# Thompson Oral History > Design, conduct, preserve, evaluate, interpret, and present oral-history research using Paul Thompson's The Voice of the Past. Use when recovering excluded voices, building community or life-history projects, selecting narrators, conducting recorded interviews, assessing memory and representativeness, creating archives and transcripts, or integrating oral testimony with documentary evidence. Treat testimony as situated, co-produced evidence rather than either literal fact or disposable anecdote. - Skill: `ilog3/thompson-oral-history` (Agent Skill, multi-file: 12 files) - Install (CLI): `npx skillmds@latest add ilog3/thompson-oral-history` - Raw SKILL.md: https://api.skillmd.com/api/skills/ilog3/thompson-oral-history/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: ilog3 (https://skillmd.com/u/ilog3) - Updated: 2026-09-22 - Page: https://skillmd.com/skills/ilog3/thompson-oral-history --- # Thompson Oral History Use Paul Thompson's oral-history framework to turn remembered experience into ethically collected, critically tested, publicly usable historical evidence. ## Core stance - Begin with purpose: ask whose past is being reconstructed, by whose authority, and for which community or audience. - Treat oral testimony as both evidence about events and evidence about meaning, identity, silence, and later reinterpretation. - Do not rank written records as inherently superior. Criticize every source according to how it was produced, preserved, selected, and used. - Preserve the narrator's life context while allowing thematic comparison across interviews. - Keep collection, evaluation, interpretation, archiving, and presentation connected from the start. ## Workflow 1. **Frame the historical intervention.** State what established account, archival silence, social group, or community need the project addresses. Read [chapters/ch01-history-community-and-oral-evidence.md](chapters/ch01-history-community-and-oral-evidence.md). 2. **Define the evidence claim.** Separate event reconstruction, lived experience, retrospective meaning, collective memory, and narrative form. Read [chapters/ch02-achievements-and-evidence.md](chapters/ch02-achievements-and-evidence.md). 3. **Design the project.** Bound the topic, period, place, population, outputs, resources, and participation model; prepare a topic guide and sampling frame. Read [chapters/ch04-project-design-and-sampling.md](chapters/ch04-project-design-and-sampling.md). 4. **Select narrators strategically.** Match random, quota, stratified-community, network, key-informant, or theoretical sampling to the intended inference. Never claim statistical representativeness from convenience recruitment. 5. **Establish ethical terms.** Explain recording, confidentiality, attribution, withdrawal, copyright, preservation, access, and future reuse before recording. Reconfirm permissions when outputs change. 6. **Conduct an open but prepared interview.** Invite a broad life narrative, listen without needless interruption, then probe chronology, concrete incidents, meanings, contradictions, and sources. Read [chapters/ch05-interview-practice.md](chapters/ch05-interview-practice.md). 7. **Create the evidence package.** Retain the preservation master, working copy, metadata, field memo, consent and access record, transcript or indexed summary, and links to related sources. Read [chapters/ch06-preservation-transcription-and-rights.md](chapters/ch06-preservation-transcription-and-rights.md). 8. **Criticize and triangulate.** Test internal consistency, chronology, plausibility, narrator position, interviewer influence, survival bias, and agreement or productive disagreement with other sources. 9. **Interpret without flattening.** Move between whole lives and cross-case themes; preserve contrary cases, silence, emotion, language, and narrative organization. Read [chapters/ch07-interpretation-and-presentation.md](chapters/ch07-interpretation-and-presentation.md). 10. **Report calibrated conclusions.** Distinguish what the project establishes, suggests, cannot resolve, and newly enables people to ask. ## Decision rules ### Choose the interview form - Use a **life-history interview** when sequence, turning points, identity, and relations across domains matter. - Use a **topic-focused interview** when the claim concerns a bounded event, workplace, institution, practice, or movement. - Use a **group interview** when interaction can recover shared vocabulary and mutually stimulated memories; document conformity and dominance risks. - Use **repeat interviews** for complex lives, sensitive subjects, unresolved chronology, or collaborative interpretation. ### Interpret memory - A remembered concrete practice may be more stable than a past attitude stated retrospectively. - Telescoping, compression, omission, hindsight, public narratives, and present identity can reorganize recall. - Inaccuracy about a date does not erase testimony's value for studying meaning; symbolic truth does not establish event fact. - Silence is a prompt for contextual inquiry, not automatic proof of repression. - Do not act as a therapist. Pause when distress exceeds the agreed research relationship and provide an appropriate referral route when available. ### Judge adequacy Ask four separate questions: 1. **Authenticity:** Is the recording and record what it claims to be? 2. **Accuracy:** Which factual elements survive internal and external checks? 3. **Representativeness:** What population, experience, network, or cultural position can these narrators illuminate? 4. **Meaning:** What does the way a story is told reveal that an event ledger cannot? ## Required outputs For a project plan, provide the historical purpose, evidence claims, narrator population, sampling logic, topic guide, ethics and rights plan, preservation workflow, source-criticism matrix, analysis strategy, limitations, and return-to-community plan. For an evidence analysis, use the templates in [patterns.md](patterns.md) and vocabulary in [glossary.md](glossary.md). For rapid field use, consult [cheatsheet.md](cheatsheet.md). ## Guardrails - Do not present a narrator as a passive data container; the interview is shaped by both parties. - Do not erase grammar, dialect, hesitation, or emotion without disclosing editorial choices. - Do not publish sensitive material merely because it was recorded. - Do not infer prevalence from vivid testimony or treat one exemplary life as an average life. - Do not isolate quotations from life sequence, question context, or the recording. - Do not let obsolete recording technology in the source dictate current technical practice; retain its principles with contemporary formats. ## Chapter map - Chs. 1–3: [history, community, historiography, and new fields](chapters/ch01-history-community-and-oral-evidence.md) - Ch. 4: [evidence, reliability, representativeness, and source criticism](chapters/ch02-achievements-and-evidence.md) - Ch. 5: [memory, self, emotion, language, and silence](chapters/ch03-memory-and-self.md) - Ch. 6: [project design, participation, and sampling](chapters/ch04-project-design-and-sampling.md) - Ch. 7: [interview preparation and conduct](chapters/ch05-interview-practice.md) - Ch. 8: [preservation, indexing, transcription, access, and rights](chapters/ch06-preservation-transcription-and-rights.md) - Ch. 9: [analysis, interpretation, media, and public presentation](chapters/ch07-interpretation-and-presentation.md) ## Further resources - [glossary.md](glossary.md) — key concepts. - [patterns.md](patterns.md) — reusable planning and analysis templates. - [cheatsheet.md](cheatsheet.md) — compact operational checklist. This skill is a structured methodological synthesis of Paul Thompson's *The Voice of the Past: Oral History*. Apply it critically in light of current law, ethics, consent standards, and recording technology.