Methods[edit]

The first critical step of language description is to collect data. To do this a researcher does fieldwork in a speech community of their choice, and they record samples from different speakers. The data they collect often comes from different kind of speech genres that include narratives, daily conversations, poetry, songs and many others.^[12] While speech that comes naturally is preferred, researchers use elicitation, by asking speakers for translations, grammar rules, pronunciation, or by testing sentences using substitution frames. Substitution frames are pre-made sentences put together by the researcher that are like fill in the blanks. They do this with nouns and verbs to see how the structure of the sentence might change or how the noun and verb might change in structure.^[12]


There are different types of elicitation used in the fieldwork for linguistic description. These include schedule controlled elicitation, and analysis controlled elicitation, each with their own sub branches. Schedule controlled elicitation is when the researcher has a questionnaire of material to elicit to individuals and asks the questions in a certain order according to a schedule.^[13] These types of schedules and questionnaires usually focus on language families, and are typically flexible and are able to be changed if need be. The other type of elicitation is analysis controlled elicitation which is elicitation that is not under a schedule.^[13] The analysis of the language here in fact controls the elicitation. There are many sub types of analysis controlled elicitation, such as target language interrogation elicitation, stimulus driven elicitation, and many other types of elicitation.^[13] Target language interrogation elicitation is when the researcher asks individuals questions in the target language, and the researcher records all the different answers from all the individuals and compares them. Stimulus driven elicitation is when a researcher provides pictures, objects or video clips to the language speakers and asks them to describe the items presented to them.^[13] These types of elicitation help the researcher build a vocabulary, and basic grammatical structures.


This process is long and tedious and spans over several years. This long process ends with a corpus, which is a body of reference materials, that can be used to test hypothesis regarding the language in question.