Timbuktu Guidelines on Oral Testimony
Developed by ACORD in Timbuktu, Mali 7 -11 May 2000

Introduction
Oral testimony is a powerful tool for gaining an insight into people’s personal experiences and perceptions. It often uncovers aspects of people's lives which otherwise remain hidden, and allows us to understand subjective positions. If done well it can help us to cut across barriers such as wealth, gender, class, ethnicity etc. People's perceptions (even if we disagree with them) help us to understand the situation and their actions better.

Oral testimony is also a useful tool for addressing politically and socially sensitive topics, which are difficult to enquire into using other methods. As a method it is complementary to other research methods such as PRA and quantitative surveys, and the findings from oral testimony are often made more powerful if combined with results from other methods.

The following guidelines are based on the experiences of ACORD staff who came together in a workshop in Timbuktu from 7 - 11 May 2000. All had used oral testimony in their research work, often to address politically sensitive topics. The guidelines are in three main sections: preparation, conduct of research, analysis and usage of research. They were drawn up on the basis of discussions about the use of oral testimony to research issues relating to Gender and Conflict, but they are applicable to the use of oral testimony more generally.

• Be multidisciplinary
• Have a range of language skills
• Take into account cultural and gender aspects as well as context. It is useful to have a range of different types of people represented on the team and to establish which specific characteristics are most appropriate for interviewing particular categories of interviewees. However, it is also important not to make too many assumptions based on stereotypes about who will talk to whom; the individual's character and personal qualities can be as or more important than their social identity in determining their success as an interviewer.

1.2    Qualities of the team
Most importantly, whatever their social identity, whether in terms of disciplinary background, gender, age, race, ethnicity, etc. researchers need to:

• Be able to listen without making preconceived assumptions and without passing judgement on the respondent
• Be patient, sympathetic and a good listener
• Be willing to give their own oral testimonies
• Be able to understand and extract ideas, even if they are implicit rather than stated explicitly by the respondent
• Be good communicators who can overcome communication blocks
• Have knowledge of the local context/situation and to respect other peoples' cultures
• Be willing to learn

The team should be trained in oral testimony methods before conducting fieldwork, and should cover all the guidelines set out in this document. Training is necessary to ensure that interviewers have the necessary degree of professionalism to work on sensitive issues.

The training process also goes hand in hand with identifying the broad issues to be covered in collecting oral testimonies. It should involve developing the team's understanding of the nature of the community, the nature of the problems or situation, the team's own engagement in the community and the language spoken.

• The nature of the community, or community profile, can be understood through examining secondary data, drama, local songs, etc., or through conducting PRA exercises

• The team can carry out preliminary oral testimonies with each other and other staff members, possibly starting with the team leader. This helps to

• develop an understanding of what it is like to give an oral testimony

• develop listening skills

• understand an individual staff member's involvement in the issues being addressed in the research. These may be analysed jointly by the team members (which requires a serious commitment to confidentiality)

• identify particular phrases, expressions and words that are used to discuss certain issues. This is particularly important where the research theme touches on sensitive or taboo topics. Asking the questions in the right way is as important as asking the right questions

• All members of the team should be involved in the formulation/identification of objectives, issues and methods. As they work through the issues that are being researched (e.g. gender and conflict), they need to challenge their own assumptions and have a common understanding of the concepts and aims of the project. The outcome of this joint preparatory work should be a set of themes and reference questions that the team to focus on (not to be taken out as a questionnaire but as a conceptual guide). In the preparatory phase, the whole team should try to identify which methods will be most appropriate for the different types of information.

• An important part of the training process is to understand that men, women and different cultures often express emotions in different ways, and to identify some of the norms of expression in the field sites.

• People very often mask their real emotions and present them indirectly. It is important to learn how to 'decode' these indirect forms of expression. Members of a particular community are sometimes, but not always, the best people to do this. Again, working in a team helps to overcome these issues.

• Training must be ongoing. A very useful tool is to ensure that interviews are transcribed rapidly and systematically and then shared with fellow team-members that can comment on the interview and suggest further questions that could have been asked.

• It may be useful to develop specific interviewing skills for specific categories of people. For example, in some places, it may be more difficult to interview men about their personal experiences because they have been socialised not to express their emotions

• Some flexibility should be built into an oral testimony research timetable. For example, a person may not always be available for interview at the scheduled time. It is also very useful to make several visits to collect one person's testimony. People are often much more open in a second or third discussion than in the first. In oral testimony, QUALITY is more important than QUANTITY.

• It may be useful to think of the research in phases and to use different methods. For example, oral testimonies could be used to develop a more quantitative survey or to get a more in-depth illumination of quantitative survey findings.

• It is usually important that the team carries out preparatory consultation with the authorities and with ordinary members of the community in the area where it wishes to carry out its work. This would usually cover the objectives, areas of interest and possible uses of the materials. In some places it is essential to obtain a letter of authorisation from the local authorities who wish to have a clear understanding of what the researchers are trying to achieve and how the results will be used.

• It is important to plan from the beginning how the testimonies and the other forms of data will be stored. Particularly in the case of testimonies which may contain very personal and sensitive information, DATA MUST BE STORED IN A SECURE WAY, and only the researchers should have access to it.

• As a broad guideline, whichever method is used, allow equal time for the preparation phase, the data collection phase, the analysis and writing up phase and the feedback, dialogue and dissemination phase.

1.5     Targeting/sampling

• Consider whether particular categories need to be identified. Sometimes it is sufficient to identify people involved in a particular type of activity, e.g. ex-combatants, prostitutes, prisoner. Often though, it is important to narrow the focus to other criteria such as age, gender, economic status, and ethnicity e.g. young male prostitute, middle-aged female ex-combatant.

• Try to collect testimonies from all actors who are known to be relevant to the issues being researched.

• People's experiences are often significantly different from one area to another. Discuss whether it is important to have respondents from a range of different locations.

• A key criterion is people's willingness to participate. Always try to start interviewing people with whom there is a relationship of trust. This may include people who already have close relationships with ACORD and its staff

• Sample size is dependent on how many different categories are seen as being directly relevant to the research issue, and on the overall budget and timescale of the project. A sample of ten people per category may be the minimum necessary to ensure representativity.

1.6     Potential pitfalls of using oral testimonies

• The method does not automatically allow taboo topics to be uncovered - success depends on all the skills and qualities outlined above
• It is very difficult to record oral testimonies if the respondent does not allow the use of tape recorders or note-taking
• Because only a small number of interviews are likely to be conducted in any given place, the respondents may become the objects of unwelcome attention. Other people may suspect that the respondents have benefited directly in some way or that they have given away secret information which could be used against them
• It may be difficult for ACORD to address the issues which emerge from oral testimony work
• As in all research work, people may have raised expectations of some concrete benefit, even if it is made clear that nothing is being offered in that regard
• Some findings may be too sensitive for public dissemination
• If used alone, oral testimony findings may not be seen as representative

2. CONDUCT OF RESEARCH
2.1    Risk/security
In many of the situations where ACORD does research, it is essential to be aware of the security situation - both physical and political security is important. When planning and carrying out such research, it is important to ensure that:

• testimonies and related documents/data are kept in secure places

• administrative, moral and political protection is obtained from the organisation and from relevant leadership figures in the area
• anonymity of respondents is maintained, and data is used with great discretion

• ACORD is prepared to meet legal defence costs

• ACORD furnishes interviewers with letters of identification and a statement of support. This can normally be drawn up at programme level, but in some instances it could come from the country office or the secretariat

• If computers are being used to type up and store transcriptions, there should be good computer security e.g. password systems to access the data

2.2     Access

• Where necessary and possible, authorities should be informed about what is being done - if necessary an authorisation letter can be obtained from them. In some field situations, it may be useful to start doing oral testimonies with local leaders, though in other situations, it may be a poor strategy. Sometimes, the main issue is to get local leaders to commit themselves to supporting and protecting interviewees, rather than having them interviewed.

• As stated above, interviews should only be done with willing respondents - there is no value in trying to pressurise somebody

• If it makes it easier to gain access, build oral testimony collection into other work

• In strict security areas consider using local partner organisations to do the oral testimony (train them). However, there are also risks: do you know where they stand politically? Who they network with? Are they stronger or more vulnerable than ACORD? All of these are context specific

2.3     Trust and confidentiality
Once access has been established:

• Establish a relationship (if not already existing) with the respondent and organise an interview time and place that suits them

• Assure interviewees that their testimonies are confidential. The interview team must maintain confidentiality at all times – not tell people who, where, when, etc. This is essential to the integrity and credibility of the team. If testimonies are disseminated, the interviewee's identity should remains anonymous

• Consider showing a letter of support from ACORD if it helps to gain the respondents' confidence

• Explain the purpose and use of the oral testimonies. Explain sharing risk and explore some of the possible benefits to the respondent. These might include giving a voice to the interviewee, solidarity, opportunity to express difficult issues and experiences, more appropriate interventions (ACORD). NEVER try to persuade a potential respondent to give an oral testimony by making false promises for what they will receive in return

• Explain why you would prefer to make a tape recording i.e. writing reduces listening, things might be missed, needs full attention etc. but respect the interviewee's wishes if they do not want their testimony to be taped or written down. Don't be rigid in insisting on using tape-recorders - it is not always appropriate. It may be useful to tell people that the tape-recorder can be switched off at any point and that recording limits the possibility of distorting what they say

• Establish common ground; it often helps to share experiences at initial and subsequent meetings. When interviewees express strong emotions, do not be embarrassed to share them. However, do not continue to do so if the interviewee's emotions or environment are not conducive. Sympathise and share experiences, if it is appropriate, and arrange to continue with the oral testimony at another time.

• Dress appropriately for the interview environment and in a way that puts you and your respondent at ease; being over- or under-dressed can make you or your respondent (or both) quite uncomfortable and create a barrier to communication

• Be aware of body language. In some areas, respondents are more at ease if they are sitting close to the interviewer and sharing food and drink, while in other areas, respondents prefer to sit far apart. In other words, 'when in Rome, do as the Romans do'.
  
• Recognise cultures and traditions that help integration. For example, in certain places, offering a coffee or buying a beer are normal forms of hospitality and bringing a small gift of sugar or salt is the normal behaviour for a guest.

• Sometimes, it is easier for two members of the research team to interview a respondent together. In other instances, it is easier to interview two respondents together rather than individually.

Once the discussion has started, it is important to:

• Ask open-ended questions and allow people to follow their own train of thought rather than trying to direct them. If what the respondent is saying is irrelevant, do not try to stop them. Try to use questions later on in the interview to bring them back to the relevant issue. This allows new and important issues to be discovered which might not have been thought about when planning the research.

• Avoid feeding assumptions into the discussion - try to only play a catalytic role and avoid, at all cost, passing judgements on the interviewee. Also be careful about revealing personal political views which could influence the testimony

• Follow through statements. Probe the interviewee and try to get details

• Respond to indications that something may be a perception – probe to understand why the respondent has that perception

• Allow silences, but don't let them be too long

• After finishing the interview, make notes about the interview itself - the person's behaviour, events, feelings, reactions, interruptions. Attach these notes to the transcription

Once the testimony collection is completed, the team should work to identify:

• common features
• differences between, and within, the categories of respondents
• surprises and major problems
• quantitative data
• supporting data
• timelines
• comparisons of oral testimonies across the programmes

• Preliminary results can be discussed with the relevant groups and further analysed with them. Where possible, the interviewer should give feedback to the interviewee, seek confirmation of the findings, give copies of any photos taken etc.

• Testimonies which are not immediately relevant can be filed for future reference, bearing in mind the Tamashek motto: ‘everything you keep, you will need one day, except bitterness’

• it is inevitable that what the interviewer remembers and how she or he interprets it, will involve a degree of subjectivity

• if the interviewers are disturbed by the material covered in the testimonies, it may be necessary to provide them with counselling

3.2     Utilisation

• Advocacy (consensus on use). If you cannot risk using oral testimonies for advocacy locally or internationally, consider linking human rights organisations directly to the respondents

• Influence on programming; develop projects to address issues raised and reorganise what you are already doing to respond to the real needs

• Publication and dissemination

• Promoting dialogue within and between groups and institutions in the community as well as at national/international levels. This may help in catalysing community-led initiatives. In Gulu, for example, the use of oral testimony led to the creation of a local NGO which has been active in a number of peace initiatives at different levels