Journal of Ethnopharmacology 100 (2005) 30–36

Perspective paper

Questionnaire surveys: Methodological and epistemological problemsfor field-based ethnopharmacologists

Sarah Edwards∗, Sabine Nebel∗, Michael Heinrich∗

Centre for Pharmacognosy and Phytotherapy, The School of Pharmacy, University of London, 29-39 Brunswick Sq., London WC1N 1AX, UK

Accepted 18 May 2005Available online 11 July 2005

This paper is dedicated to S. Edwards’ Kugu-Uwanh adopted father, J. Ngallametta, who passed away during its preparation. He was a respected Songman,and one of the last keepers of much traditional Wik and Kugu knowledge. He will be greatly missed.

Abstract

The classical scientific approach is empirical. One of the favoured means of gathering quantitative data in the health and social sciences,including ethnopharmacology and medical ethnobotany, is by use of questionnaires. However, while there are numerous published articlesd r of culturalc y the use ofq ros and conso©

K

1

i(fpceEhlsppam

(

har-iencethat

nter-chal-inglyarchitativentificen-

itingtion-that

tionsl dif-l

pre-ed in

0d

iscussing the importance of questionnaire content, the fact that questionnaires themselves may be inappropriate in a numbeontexts, even where literacy is not a factor, is usually ignored. In this paper, the authors will address the main issues posed buestionnaire surveys, using case studies based on their own personal experiences as ethnopharmacologists ‘in the field’. The pf qualitative and quantitative research and the use of alternative means to elicit quantitative data will be discussed.2005 Published by Elsevier Ireland Ltd.

eywords:Ethnopharmacological methods; Questionnaires; Traditional knowledge; Australia; Italy; Mexico

. Introduction

The interdisciplinary nature of ethnopharmacology, whichs very important in the study of traditional medicineLipp, 1989; Heinrich et al., 2004), encompasses aspectsrom a number of diverse subject areas, including anthro-ology, botany, clinical research, pharmacognosy, pharma-ology, natural product chemistry and toxicology. How-ver, a concern raised by the International Society ofthnopharmacology (see editor’s note on the website,ttp://www.ethnopharmacology.org/) is that ethnopharmaco-

ogical studies are predominantly focused towards the naturalciences with relatively few undertaken from a social scienceerspective. One reason for this is possibly because anthro-ology is often perceived pejoratively as a ‘soft science’, andlthough some anthropologists apply bioscientific methods,any more use qualitative methods apparently lacking in sci-

∗ Corresponding authors. Fax: +44 20 7753 5909.E-mail address:phyto@ulsop.ac.uk

S. Edwards, S. Nebel, M. Heinrich).

entific rigour. Given that the majority of academic ethnopmacological studies are undertaken through natural scdepartments or institutions, it is therefore not unexpecteda social or cultural emphasis is frequently neglected.

Field-based ethnopharmacologists who work at the iface between the natural and social sciences often facelenges in how to marry these two approaches. Increasin recent years in the related disciplines of health reseand ethnobotany, there has been a push to use quantmethods in order to generate data that stands up to sciescrutiny. This has been driven in particular by funding agcies. One of the most commonly used methods of elicquantitative data in the social sciences is through quesnaire surveys. However, while it has been recognisedthere is a need for culturally appropriate ways to ask quesin these surveys, little has been written about the potentiaficulties of usingquestionnaires themselvesin some culturacontexts.

In this paper, three individual case studies will besented, discussing the problems that the authors faccontrasting cultural and environmental settings.

378-8741/$ – see front matter © 2005 Published by Elsevier Ireland Ltd.oi:10.1016/j.jep.2005.05.026

S. Edwards et al. / Journal of Ethnopharmacology 100 (2005) 30–36 31

2. The Wik experience (Australia)

2.1. Field site

Collaborative ethnopharmacological research took placeat the invitation of the Wik peoples of Aurukun, an ex-mission, and currently the largest Aboriginal communityin Cape York Peninsula, far north Queensland (Australia)between 2001 and 2003. Aurukun is classified as being ‘veryremote’ according to the Australian Bureau of Statistics, andis situated on the western side of the Cape, on the Gulf ofCarpentaria. Overland, access to Aurukun is restricted tofour-wheeldrive vehicles via a dirt road, which is cut off fora few months each year during the monsoonal ‘wet’ seasonwhen travel is restricted to air or sea. The habitats aroundAurukun are diverse, consisting of mangroves and vast wet-lands (both saltwater and freshwater), Melaleuca swamps,dense forest dominated by Eucalyptus species, open savannaand riverine gallery forest. Bauxite mining has been proposedfor a large area in the interior, mainly eucalypt forest, butaside from some pastoralism for cattle established by themission (which ended after an outbreak of TB in the 1970s),to date there has been very little non-traditional use of theland.

Traditionally hunter-gatherers, ancient land managementpractises, such as burning, are still undertaken by Wik tra-d rre-s wn.T susi d asi lfi teredw andi cen-s te itw ospi-t ousp uis-t herR hem dis-t ces.T n isa ct, oro ofWf r-i hisp

al-e ts inA ncila lop-m tiallyp sicw

2.2. Health and healing

Many Wik people live in overcrowded, inadequate housingand a disproportionate number suffer from health and socialproblems, with alcoholism, domestic violence and suicidebeing commonplace in the community. Statistics show thatthere is a huge health disparity between Australia’s indige-nous and non-indigenous populations (Australian Bureau ofStatistics, 2002). Some of these can be attributable to eco-nomic deprivation, lack of access to suitable health careand poor diet (type II diabetes is prevalent in Aurukun, cor-responding to the transition from a traditional diet rich inprotein and plants to a ‘store food’ diet dominated by refinedcarbohydrates).

Compared with other communities, Aurukun is fortunatein having its own local Health Clinic, run by the governmen-tal organisation Queensland Health, which provides primaryhealth care to Aurukun’s residents. The Clinic, however, isnot always the first port of call for ill people in the commu-nity. Some sicknesses, especially those of a serious nature,are considered by Wik people to have a spiritual or sorcery(puriy–puriy) causation, and traditional healers (or witchdoc-tors, known asNoyan) are in these cases called upon to bothdiagnose and treat the sick. The tools employed by a WikNoyanare a piece of string (tied around the patient) and acup of water into which objects, such as ‘bad blood’, smalls spata u-a arms tt m toa es ofa ers orec

lledo ingt ature,sl lderm andw

2

s oft calf althW notk

e(

ed“ lant,

itional landowners, although the extent to which this coponds to fire regimes prior to mission days is little knohe population of Aurukun according to the last cen

n 2001 was 1034, of which 907 persons were classendigenous (Australian Bureau of Statistics, 2001). The actuagure is somewhat higher, according to numbers regisith the local Health Clinic (personal communication),

s around 1300 of which 1200 are indigenous. (Theus did not take into account people away on the daas taken, such as those staying on outstations, in h

al, prison or visiting relatives elsewhere.) The indigenopulation of Aurukun consists predominantly of the ling

ically related Wik and Kugu groups from south of the Arciver, and the Wik-Way people from north of the Archer. Tajority also belong to one of five spirit clans, each with

inctive ceremonial body paint markings, songs and danhe number of Aboriginal language groups in Aurukubout 15, but most of these languages are now extinn the verge of extinction, with the notable exceptionik-Ngathan and Wik-Mungkan, which remains thelinguarancaof the community. (The term ‘Wik’ will be used genecally to refer to all indigenous peoples of Aurukun in taper.)

Although there is a thriving arts centre for local tnted artists, there are few other employment prospecurukun, apart from a handful of jobs in the Shire Cound Health Clinic. There are also limited community deveent employment project (CDEP) places, which essenrovide ‘work for the dole’ and pay little more than baelfare.

tones or pieces of wire ‘extracted’ from the patient ares verification of apuriy–puriy diagnosis. Healing is uslly done by the laying on of hands and rubbing underweat on the patient. Occasionally, aNoyanis requested ahe Clinic, and clinic staff are generally amenable for thettend to their clients, especially having noted that in caspsychological or mental nature aNoyanoften has a great

uccess with their treatment than clinic staff do with monventional ones.

Phytotherapeutic remedies (or ‘bush medicine’, capar’ in Wik-Mungkan) are used in the Wik domestic sett

o treat illnesses considered to be of a less serious nuch as colds and skin infections. While theNoyanis also

ikely to be knowledgeable about plants, it is usually an oember of the patient’s family who treats a sick person,ithin their home, rather than the clinic.

.3. Research project

A major concern expressed by Wik people was losheir traditional knowledge, including knowledge of loood plants and bush medicines. A senior Aboriginal He

orker stated that young people in Aurukun today didnow much aboutopar’:

“ . . .they forget culture - but old people still have cultur”.Doris Poonkamelya Clinic Interview Notes, 2003.)

Similarly, an elderly Wik-Alkan Traditional Owner stat. . .my parents taught me the name of every tree, every p

32 S. Edwards et al. / Journal of Ethnopharmacology 100 (2005) 30–36

everyfish. . .in20years thiswill all be forgotten.Youngpeopletoday prefer to live in the busy world”.

The project was established to help address these concerns,by developing an ethnobiology database that could be usedby local community members in the school, land and seamanagement centre and clinic. The function of the databaseis to act both as a repository of local traditional knowledge ofplant foods and phytotherapies, and also to integrate currentscientific data (where it exists) with this local knowledge,giving Wik people access to the data through their local namesfor plant species. Database development and data collectionwas undertaken by Sarah Edwards (SE).

While lifestyle changes and loss of languages no doubtcontribute to loss of knowledge, it was also apparent that bushmedicines (at least a few popular ones) were still being widelyused in Aurukun, perhaps concurrently with prescribed phar-maceuticals from the clinic. As this could potentially posehealth issues due to possible drug interactions, it was felt ofinterest to find out the extent of this practice. Most clinicstaff interviewed did not enquire of their patients if they wereusing bush medicines. As one clinician explained, being givena Wik-Mungkan name for a plant would be meaningless tohim, and he therefore would not be able to make a ratio-nal assessment of its potential benefits or risks. Occasionally,however, patients voluntarily told the clinic staff that they hadused bush medicines.

2

ths itteeg findo ines.T fiden-t ionsd ablew , itw d filli

ushm kinga turedi outt sses,a bushm Wikp lantsa dicalc bies,e ver”.A r toi ionsw

s thel thea thin

last year” or “longer-term (more than a year ago)”. The rea-son for this was because local Wik concepts of time differconsiderably from the European, ‘linear’ concept, making itharder for Wik people to estimate when particular events hadtaken place. In fact, the Wik-Mungkan word corresponding to‘time’ (Aak), is the same for ‘country’ or ‘place’, time beingsituational and relative in Wik terms.

However, even given the implementation of cultural con-straints in questionnaire design it was a resounding disaster.Very few patients agreed to participate with the question-naire survey with clinic staff often experiencing an elementof hostility from patients they asked, including from thosewho moments before had been talking quite freely about bushmedicines.

Clinic staff discussed the issue with the Aboriginal HealthWorkers and a few community Elders to try and establish whythe questionnaire returns were so low. There were a numberof reasons given, but primarily the majority of Wik peopledid not want to admit to using bush medicine. Most thoughtthat it might offend the clinic if staff knew that their patientsalso usedopar’. Also, bush medicine is ‘secret business’, andpeople did not want to disclose this information to clinic staff.A proposal to translate the questionnaire into Wik-Mungkanand for the Aboriginal Health Workers to undertake the sur-vey was also felt unlikely to work, since all of these issueswould still apply. Although SE had built up a good relation-s intoa om-m ires whoa y!

otherf rk inA plea e’ ift ial-i arch,e and2 inalp1 witha mat-t

3

3

pedf nts.T andh eo gcp

.4. Questionnaire

In conjunction with Aurukun’s Health Clinic, and wiupport from Queensland Health (whose Ethics Commave their approval), a questionnaire was developed tout how many of the clinic’s clients also used bush mediche questionnaire was developed to be anonymous, con

ial and non-compulsory, with the stipulation that questid not have to be answered if the patient felt uncomfortith them. Due to low levels of literacy in the communityas decided that clinic staff would ask the questions an

n the questionnaires for their clients.The questions avoided asking about specific b

edicines, so as to maintain cultural protocols of not asbout traditional, sacred knowledge. Instead, it was struc

n a way using tick boxes to elicit information including abhe current extent of use of bush medicines to treat illnend which illness categories people had treated usingedicines. The illness categories were those given byeople in the course of recording information about the pnd their uses, and did not follow any standard biomeategorisation, for example, “skin problems (sores, scatc.)”, “stomach problems/diarrhoea” and “headache/fedraft questionnaire had been given to clinic staff prio

mplementation for comments, and a few minor alteratere made.One of the changes was to the question “when wa

ast time you used bush medicine to treat an illness”, withvailable tick boxes being reduced to two, “Recently, wi

hip with Wik colleagues (and in fact had been ‘adopted’family), when she also tried asking a few people in the cunity if they would like to participate in the questionna

urvey, they all gave a negative response, except for onegreed to participate only on being paid a sum of mone

In addition to these reasons expressed, there areundamental reasons why questionnaires may not woustralian Aboriginal contexts: culturally Aboriginal peore reluctant to expose themselves, and often feel ‘sham

hey become the focus of attention. After years of colonsm and being marginalized, to be the subject of resespecially given the legacy of colonialism and its 19th0th century anthropologists, who often regarded Aborigeoples little better than museum or zoo specimens (Griffiths,996), it is quite understandable that a request to complyquestionnaire by discussing what is viewed as a private

er, would result in a negative reaction.

. The Mixe experience (Mexico)

.1. Field site

The Mexican Mixe project questionnaires were develoor use in a large rural community of about 6000 inhabitahe land of the Mixe extends mostly through the coolumid mountains of theSierra de Juarezin the Mexican statf Oaxaca.San Juan Guichicoviis the only Mixe-speakinommunity in the subtropicalIstmo de Tehuantepec. It is therincipal community (cabecera) in a subdistrict (municipio)

S. Edwards et al. / Journal of Ethnopharmacology 100 (2005) 30–36 33

of the same name. In 1980, themunicipiohad 20,000 inhabi-tants, while thecabecera5500–6500 (Heinrich et al., 1992).Seventy-five percent of the population in thecabeceraareconsidered to be bilingual. A minute fraction of the pop-ulation can only speak Spanish. The economy is based onsubsistence agriculture (mainly maize) and on the produc-tion of coffee and citrus fruit. Another relevant commercialproduct ishuipilesof the Tehuantepec style. They are pro-duced by women and men of the community and are usuallysold to Isthmus Zapotec merchant women, who resell them innumerous communities of the Oaxacan part of the Isthmus,and in many other regions of Mexico. No detailed monographon the lowland Mixe is available. Two cultural aspects thathave been dealt with in detail are the ritual calendar, which isstill used in some parts of themunicipio(Weitlaner-Johnsonand Weitlaner, 1963) and, in some adjoiningmunicipios, therelationship of religious ritual to medical concepts (Heinrich,1994and references therein).

3.2. Health and healing

There are at least 15 different types of healers (pa’amiixyp’) known in the community. The largest groups are“specialists in home remedies”. Other important groups aremidwives, chupadores(healers that suck out an illness),prayer makers (rezadores), spiritists (espiritistas) and spiritu-a ishta inc sons( ternm , var-i ncesb trave xpe-r tans,w Overt actis-i sentb canod ista( rav-e ce onh le ont ed toi ed in1 ths).

3

cteda stion-n ky,n con-s elec-t t 13

months as part of a study on the medicinal plant use of theLowland Mixe, by Michael Heinrich in collaboration withNereyda Antonio B., a biologist, and with various healerswhose medico-botanical knowledge was recorded (Heinrich,1989; Heinrich et al., 1992). This study was developed priorto the comparative tools produced for subsequent research inlowland Mexico (Heinrich et al., 1998).

3.4. Questionnaire

The initial goal for the questionnaire element was to seehow the knowledge about medicinal plants is distributedbetween the genders, in age groups, along the cultural contin-uum from conservative Mixe speakers to mestisized membersof the community and according to socio-economic status.Overall 182 inhabitants were interviewed. Only part of thisinformation was published (Heinrich, 1994), in part becauseof lack of time, but also because of the complexity of the anal-ysis required. One crucial outcome was the identification ofpreferred forms of treatment of eight popularly recognisedillnesses.

From a methodological perspective, an approach like theone used in San Juan Guichicovi raises a whole series ofquestions. A vivid recollection is of the response of one ofthe key informants, on being told about the new element ofthe project:“Why, haven’t I told you everything?”

therh ce alli haelH wnt boutt alsop , thei

cteda thei the‘ eresti oalso beene res-s for-m dedf der-t t ther icinalp mali andh

ling.S -q wouldh e firstp wasc ren,a ru-

lists (espiritualistas). Considerable differences distinguhe various groups of healers (Heinrich, 1994; Heinrich etl., 1998). Recently, the number of those offering helpase of illness has grown to include travelling salespermostly Mixe-speakers with minimal experience in Wesedicine), assistants to the local Roman Catholic priest

ous protestant groups and trained nurses. The differeetween these health care providers are enormous. Thelling salespersons and other individuals with minimal eience in Western medicine is best considered as charlahile others have a solid background in medical therapy.

he past 10 years, there have always been one to four prng medical doctors present. Some are medical doctorsy Mexican government agencies, such as Institute Mexiel Seguro Social (IMSS) and Instituto Nacional IndiginINI). Some of the “specialists in home remedies” and tlling salespersons sell pharmaceuticals and give adviow to administer them. No quantitative data are availab

he importance of these forms of medication as opposndigenous phytotherapy. Data in this area were collect985 and 1986, and later in several short trips (1–2 mon

.3. Research project

Looking back at research which was mostly condubout 20 years ago, and the experience of using a queaire in a Lowland Mixe community in Oaxaca is tricot only because the scientific framework has changediderably, but also because one’s memory is certainly sive and subjective. Fieldwork was undertaken for abou

-

A similar reaction was experienced on contacting oealers in the community and in surrounding areas. Sin

nterviews were conducted by Nereyda Antonio and Miceinrich (together with a translator), who were both kno

o work with healers, the interviewees were puzzled ahe interviewers’ interest in their concepts. They wereuzzled about the house-to-house survey. Frequently

nterviewers were asked:“Why do you do this?”A few people seemed uncomfortable in being conta

t home, but the vast majority simply seemed to enjoynterest in their knowledge. This novel element to whatoutsiders’ were doing renewed the already intense intn the researchers’ activities. However, explaining the gf such a sociological tool to someone who has neverxposed to such interviews is problematic, and the impion gained over the whole of the period was that the inants found it difficult to understand what this was inten

or. This was in marked contrast to the other work unaken in the study, where people clearly understood thaesearchers were recording the healers’ usage of medlants. The questionnaire also resulted in a lot of infor

nformation about the research project being discussedelped in disseminating information about it.

More problematic remained the issue of random samphort-term migration to outlyingranchos, for example, is freuent, and clearly the sample was not as random as oneave expected. The researchers decided to interview therson in every third house in a street, but the samplelearly biased to the elderly, to women with young childnd to people who were sick or chronically ill. Another c

34 S. Edwards et al. / Journal of Ethnopharmacology 100 (2005) 30–36

cial issue is the number of interviews conducted. Althoughfortunate that during several weeks of intensive work it waspossible to interview a relatively large number of informants,in other cases the sample size was far too small for meaning-ful comparisons.Martınez et al., 1998, for example, lookedat forms of home treatment by mothers in different commu-nities of rural Mexico with the goal of understanding whatform of treatment is administered by the primary carers ofyoung children, their mothers. The communities selectedwere from different regions of Mexico and included Ameri-can Indian as well as Mestizo (Spanish speaking Mexicans ofmixed descent) populations. Mothers of young children wereinterviewed about the type of home treatment their childrenreceived during their last episode of diarrhoea. The samplesize in the communities was in the range of 25–41, making astatistical analysis problematic.

4. The Graecanic experience (Italy)

4.1. Field site

In the Graecanic project, questionnaires were developedto be used in Galliciano in southern Italy. The Graecanic areais part of the cultural and linguistic heritage of the MagnaGraecia (Greater Greece, 8th centuryb.c.) and the history oft

ternM , theG tainso andt theG opu-lc TheeC eaksC geo llyG ), ofw thec sub-s ilable,id hichf ande o theg com-m

4

d,b th ofp wildf er of

cuisines, partly because such food plants are commonly col-lected by women and children, but also because they falloutside the limits of food science and nutritional anthropol-ogy. Wild food plants are a supplementary nutrition sourcecontributing to a healthier diet, but they also represent a poten-tial source of income for rural communities. Wild greensare calledta chorta in Grecanicoand are consumed ona weekly to daily basis when they are in season. Elderlywomen, mainly housewives, are the main keepers of tradi-tional food knowledge, while men play an important role ingathering the plants and fungi that grow far away from thevillage.

4.3. Research project

This project was established to evaluate traditional wildfood plants and attributed health significance in the Mediter-ranean. Changes in lifestyle result in the risk of losing thisvaluable traditional food knowledge and with it a healthy andbalanced diet. Only a few of the younger generation, who areno longer fluent speakers of the ancient languageGrecanico,know the culturally most important wild edible plant species.Evidence from the field suggests that the loss of traditionalGrecanicplant names risks leading to a decrease in the num-ber and variety of usedchorta (Nebel, 2005).

4

ont n tos ped.I althm latedw lantd holdsw thoser d/ore Thise n andp inter-g 3 byS

oft aringf therm ity isv pledh nowl-e they sb bitr rag-i thep sizei t bep

he later Byzantine Empire in southern Italy.During the period of the Magna Graecia, the whole eas

editerranean was dominated by Greek culture. Todayraecanic area has receded into the Aspromonte Mounf the southern tip of the Peninsula (Region of Calabria)

o Salento, in the Region of Puglia. The inhabitants ofraecanic area, in respect to the surrounding Italian p

ation, are characterised by their own language (Grecanico),ulture and history as an ethnic and linguistic minority.lderly inhabitants of Galliciano are bilingual (Grecanicoandalabrian), whereas the younger generation mainly spalabrian, an Italian dialect. Italian (the official languaf Italy) is only spoken at school or by officials. Officiaalliciano has around 100 inhabitants (16 householdshich around 30% are employed in agriculture, 35% inonstruction industry and service sector abroad and atantial percentage, although no precise figures are avan forestry and related occupations. Galliciano was partiallyestroyed by two serious floods in 1951 and 1971, w

orced many inhabitants to abandon the village. Floodsarthquake together with fire and epidemics have led tradual decrease of farming and pastoral activities in theunity (Condemi, 1999).

.2. Health and nutrition

Many wild plants are traditionally used not only for foout also for medicinal purposes, as nutrition and healeople are strongly interconnected. The significance of

ood plants has been underestimated for a great numb

.4. Questionnaire

As part of this study, a socio-nutritional questionnairehe consumption of traditional food plants and its relatioocial, economic and anthropological factors was develondividuals were interviewed about the perception of heaintenance, health beliefs, other health behaviours reith the consumption of generic foodstuffs and specific pietary components. Overall 22 members of eight houseere interviewed. Household members are defined as

esiding at the time of the study and normally cook anat together one meal an average of one meal daily.nabled investigation of the characteristics, consumptioerspectives of different household members, such asenerational factors. Data was collected in spring 200abine Nebel (SN).After random sampling of eight households (= 50%

he households), the main person responsible for prepood in the household was interviewed first, and any oember of the household afterwards. As the commun

ery small, most of the housewives of the randomly samouseholds had already shared their ethnobotanical kdge about wild food plants in unstructured interviewsear before. So, the question:“Why, haven’t I told you thiefore?” was very common, and made people at first aeluctant to continue the interview. However, after encoung and explaining the purpose of the interview, most ofeople were happy to continue. In general, the sample

n this project is very small and statistical analysis mighroblematic.

S. Edwards et al. / Journal of Ethnopharmacology 100 (2005) 30–36 35

Furthermore, as the interview was on several sheets ofpaper, it looked very official to the people and increasedtheir doubts.“Can our knowledge be just in figures?”askedone of the interviewees. As at first the interview was con-ducted in Italian by SN and not the local dialect, the soundof the official language of Italy reinforced the impression ofbeing something official to the people. Later the interviewwas conducted with the help of a local person from the vil-lage, to bring it closer to the people and reassure them, thatthe information gathered is entirely for this project and hasnothing to do with the authorities. This, however, raised thequestion of the influence of a local translator present at theinterview.

Another methodological problem can be the length of theinterview. In this research, the interview had several parts,including as well a Food Frequency Questionnaire (FFQ). Inthe FFQ people were asked to give the frequency and averageamount of different foods that they have consumed over thelast 12 months. As mentioned byShahar et al. (2003), a majordilemma of an FFQ is the length of the questionnaire. ShorterFFQ’s would save time and energy for the interviewee, but anabbreviated FFQ may be less sensitive to traditional foods,and thus be less accurate.

5. Conclusions

sed tg : int sam-p xtsa erse,a g aq kinge ugh toc med‘ dentsw a sur-v herse

nicalr icro-p t off altert erb tive ofa m an‘ oot-i ties,t aret it isl n-t n allt readi inlyi ked

an antagonistic reaction due to their perceived associationwith officialdom.

These experiences are in contrast to those of Gomez-Beloz, who advocates the use of questionnaires in ethnob-otany based on his experience working with the WinikinaWarao (Gomez-Beloz, 2002), which for him was successful.One crucial piece of information that he mentions is that hecompensated all his respondents, including those who endedthe interview before it was completed.Voeks and Leony(2004)used a questionnaire in Bahia, Brazil to examine thelink between medicinal plant knowledge and age, gender andsocio-economic standing and state that they were“forced tosolicit specific individuals to participate”, actively coaxingand contributing a ‘tip’ to their informants. In the first casestudy presented, the fact that a Wik man agreed to participateon condition of ‘being paid’, along with Wik culture having astrong tenet of reciprocity, indicates that questionnaires couldalso be perceived as work or transactions.

Therefore, the question that should be asked before devel-oping a questionnaire survey is: what does a questionnaireactuallymeanto a community? Does it represent official-dom to be resisted? Or is participating in a survey perceivedas work that should be compensated, or an exchange withsomething expected in return? Generally, ethnobotanical andethnopharmacological field data are more aboutknowledge,and less about practice. In non-literate societies, it is com-m husm ues-t dersr s andc ry tor tuallyu

vea clane Abo-r ataf ntificc lita-t alln sucha ill bep nob-t mple,i inr hodw fromq lcu-l plants ed byc por-t

itativea tativer o bei nec-

In these three cases presented questionnaires were uather quantitative data from different groups of people

he first case clinic patients, secondly from a randomisedle and thirdly from housewives. While the cultural contend the specific users of plants in the surveys were divll three contexts presented similar difficulties in applyinuestionnaire approach. In the Wik case, the act of asven general questions about bush medicines was enoause affront, due to the delicate nature of what is deesecret business’. In the latter two cases, some responere surprised that they were being asked to completeey when they felt that they had already told the researcverything they wanted to know.

Field-based ethnopharmacological and ethnobotaesearch is very dependent on, and sensitive to, molitics, i.e. researcher-collaborator relations. The ac

ormal surveying through use of questionnaires canhe status quoby creating a perceived imbalance in powetween researcher and interviewee. From the perspecquestionnaire respondent, the formal survey makes the

object of study’, rather than a colleague on an equal fng. In the case of the Greacenic and Mixe communihe initial use of Italian and Spanish, respectively, whichhe ‘language of officials’, compounded this effect, andikely that use of English in the Wik community also coributed to the respondents’ reluctance to participate. Ihree examples given, respondents did not comprehendly what the questionnaires were ultimately for, and certan the Wik and Greacenic communities the forms provo

o

-

on for a select few to be ‘keepers of knowledge’, taking random surveys largely irrelevant. Restricting q

ionnaires to these specialist traditional knowledge-holesults in small sample sizes, making statistical analyseomparisons virtually redundant. It is, therefore, necessaedefine in which circumstances questionnaires are acseful.

In the Aboriginal context only traditional owners hauthority to speak about the plants that grow on theirstate, due to the sacredness of knowledge: this is theiginal Law. In cases like these, collecting quantitative dor the sake of adhering to accepted perceptions of scieredibility is unlikely to generate data as valuable as quaive data accumulated through working closely with a smumber of specialists using anthropological methods,s participant-observation. Quantitative surveys may stossible in some circumstances, but other indirect and u

rusive survey approaches should be considered. For exan Lee et al.’s paper (1994)on apparent dietary intakeemote Aboriginal communities, a ‘store-turnover’ metas used. Quantitative data can also often be generatedualitative data, for example, it may be possible to ca

ate the number of separate medicinal uses of particularpecies, and the number of times they have been mentionollaborators, thus generating a guide to the relative imance of particular species (cf.Heinrich et al., 1998).

There are advantages and disadvantages to both qualnd quantitative approaches. One advantage of qualiesearch includes ‘flexibility’, thus allowing researchers tnnovative, although critics say that this may result in a

36 S. Edwards et al. / Journal of Ethnopharmacology 100 (2005) 30–36

dotal or unstructured data, a failure to test hypotheses, and atendency to bias. However, quantitative survey methods arenot necessarily value free, and often manipulate data accord-ing to preconceptions. Extrapolation from sample surveysmay ignore possible disparities that exist in the ‘real world’,and how respondents answer questions in formal contextsmay differ from responses given in a more relaxed interview.Dependence solely on quantitative methods omits to considerthe social and cultural construction of ‘variables’, which needto be comprehended to obtain a complete and valid picture(Silverman, 2001). Scientific data that is based on repeatableexperiments or random sampling is not necessarily superiorto data generated qualitatively and is based on social realities.Science, after all, is a process of inquiry, whichever methodis used. In the area of ethnopharmacology we need to focuson developing novel methods of inquiry,andon implement-ing the knowledge ethnopharmacologists gain to benefit thetraditional custodians of this knowledge.

Acknowledgements

Our field research would not have been possible withoutthe help of many people in Queensland Australia, Mexicoand southern Italy, especially the health workers, midwives,healers and specialists in medicinal plants. These persons aret e areg dgew

amef ion( sityo ir-w el-o lowT rda

R

A ous-

Australian Bureau of Statistics, 2002. National Health Survey: Aborig-inal and Torres Strait Islander Results, Australia Cat. No. 4715.0,http://www.abs.gov.au/ausstats.

Condemi, F., 1999. Galliciano. Acropoli della Magna Grecia, Laruffa,Reggio Calabria.

Gomez-Beloz, A., 2002. Plant use knowledge of the Winikina Warao:the case for questionnaires in ethnobotany. Economic Botany 56,231–241.

Griffiths, T., 1996. Hunters and Collectors: The Antiquarian Imaginationin Australia. Cambridge University Press, 430 pp.

Heinrich, M., 1989. Ethnobotanik der Tieflandmixe (Oaxaca, Mexico) undphytochemische Untersuchung vonCapraria biflora L. (Scrophular-iaceae). Dissertationes Botanicae No. 144. J. Cramer in GebriiderBorntraeger Verlagsbuchhandlung, Berlin und Stuttgart.

Heinrich, M., Rimpler, H., Antonio, B.N., 1992. Indigenous phytotherapyof gastrointestinal disorders in a Mixe lowland community. Journal ofEthnopharmacology 36, 63–80.

Heinrich, M., 1994. Herbal and symbolic medicines of the lowland Mixe(Oaxaca, Mexico): disease concepts, healer’s roles, and plant use.Anthropos 89, 73–83.

Heinrich, M., Ankli, A., Frei, B., Weimann, C., Sticher, O., 1998. Medic-inal plants in Mexico: healers’ consensus and cultural importance.Social Science and Medicine 47, 1863–1875.

Heinrich, M., Barnes, J., Gibbons, S., Williamson, E.M., 2004. Fundamen-tals of Pharmacognosy and Phytotherapy. Elsevier Science, ChurchillLivingstone, p. 52.

Lee, A.J., O’Dea, K., Mathews, J.D., 1994. Apparent dietary intake inremote aboriginal communities. Australian Journal of Public Health18, 190–197.

Lipp, F.J., 1989. Methods for ethnopharmacological field work. Journalof Ethnopharmacology 25, 139–150.

M ter-rrhea

ical

N

S evel-FFQ)urnal

S . In:25–

V edic-94–

W re cal-19,

he original keepers of ethnobotanical knowledge, and wrateful to them for their willingness to share their knowleith us.Funding and logistical support for these projects c

rom a variety of sources, including the European UnQLK1 2001–00173); the School of Pharmacy, Univerf London; the Richard Cannell Fund (ULSoP); British Aays; Aurukun Shire Council; Balkanu Cape York Devpment Corporation; James Cook University; the Thriprust (through RBG Kew); the Lawrence Memorial Awand the British Federation for Women Graduates.

eferences

ustralian Bureau of Statistics, 2001. Census of Population and Hing, 350100250 Aurukun (s) (Statistical Local Area),http://www.abs.gov.au/ausstats.

artınez, H., Ryan, G.W., Guiscafre, H., Gutierrez, G., 1998. An incultural comparison of home case management of acute diain Mexico: implications for program planners. Archives of MedResearch (Mexico, D.F.) 29, 351–360.

ebel, S., 2005. Ta chorta: Piante commestibili tradizionali di Galliciano.University of London, School of Pharmacy, London.

hahar, D., Shai, I., Vardi, H., Brener-Azrad, A., Fraser, D., 2003. Dopment of a semi-quantitative Food Frequency Questionnaire (to assess dietary intake of multiethnic populations. European Joof Epidemiology 18, 855–861.

ilverman, D., 2001. What is qualitative research? Chapter 2Interpreting Qualitative Data. Sage Publications, London, pp.41.

oeks, R.A., Leony, A., 2004. Forgetting the forest: assessing minal plant erosion in eastern Brazil. Economic Botany 58, 2306.

eitlaner-Johnson, I., Weitlaner, R.J., 1963. Nuevas versiones sobendarios mijes. Revista Mexicana de Estudios Antropologicos41–62.