Some of your data do not exist anywhere else in computer accessible form or even in print. Apply the journalist's "what, who, when, where, why, how" criteria judiciously to fill in this information. If one of your students were to visit the area two years from now, could they find the place where you collected your data? Could they find the person?
Remember that the people who give you information may not be in a place where streets or street addresses have reached. I've collected data sitting in a canoe in the middle of a jungle river in Suriname. But I could find my consultant (if he's still alive) tomorrow, because I asked for and got quite a bit about his position in his society.
Names can be a problem:
- Names may be closely guarded secrets
- Names used openly in tense political situations may lead to reprisals
- Names may change as people go through different phases of life
What is important for the study of language change is not so much the names and identities of individuals themselves, but the sectors of society whose speech those individuals represent. Record
- Age (often estimated or pegged to some datable, memorable event)
- Generation (child, unmarried young adult, childrearing generation, grandparental generation, ...
- Sex
- Office held in local society when interviewed
- Clan, lineage, caste, or other social segment that might differentiate speech
- History of contact with other languages



Published sources for collections

