PresentationPDF Available

Elicitation techniques

Authors:

Abstract

Elicitation techniques. Invited talk, Content Development Department, Rosetta Stone.
Hieber, Daniel W. 2010. Elicitation techniques. Invited talk, Content Development Department, Rosetta
Stone, Harrisonburg, VA.
1
Elicitation techniques
Daniel W. Hieber
Rosetta Stone
1. Morphological typology
a. Morphological typology affects the way you should elicit information
i. Isolating languages you ask for the word for ‘chair’, and you’ll get
chair
ii. Synthetic languages you ask for the word for ‘chair, and you’ll get
bikáá’ dah asdáhí ‘the thing for sitting up there on’
1. Always follow up ‘what does bikáá’ dah asdáhí mean?’ ‘what
does dah mean?’
b. Polysynthesis
i. Do not think of concepts as 'words'
1. Bad: "what's the word for…?"
2. Good: "how do you say…?"
ii. Offer examples in the language
1. Does it make sense to say…?
2. What about…? What does that mean? What's the difference
between this and that?
c. Talk about what constitutes a ‘word’ in the language with your language
expert
2. Do your research
a. Learn the pronunciation and use it
b. Repeat words back to the informant, and let them correct you
c. Keep your ego in check!
d. A useful community role is that of a language learner
3. Informant Background
a. The first thing you should do when working with a new consultant is get an
idea of their background
i. Gives you a chance to ask about their knowledge of the language, as
well as their language attitudes
ii. Are you dealing with someone extremely conservative regarding
language use?
iii. Breaks the ice gets them talking about their personal life a bit more,
and makes things more intimate; invites a more trusting atmosphere
b. What was the linguistic situation like where they grew up? In their
household? Who do they speak the language with, and when?
c. Language ideologies
i. Prescriptivism
ii. Social connotations and roles
2
1. Cultural considerations e.g. in some societies, men won’t
listen to the opinion of a woman
4. Establish friendships
a. Watch for personal space
b. Use food and drink!
c. Try to relate
d. Ask about religion, history, culture, local geography, local customs, etc.
e. I start every session with a couple minutes of just shooting the breeze,
finding out what’s new in their life
i. Usually this leads to interesting cultural/ethnographic questions
f. Elicit using culturally-relevant vocabulary whenever possible
i. Many informants will add little tidbits, sentences, or pieces of
information as you go
5. Literacy
a. Heath, Shirley Brice. What No Bedtime Story Means: Narrative Skills at Home
and School.
b. How well do they read/write the language? This has a significant impact on
how they see the language.
c. Cultures have different uses for literacy that affect their language ideologies
6. Culturally-specific methods of structuring an argument or discourse
a. In many cultures, to learn about the hole in the roof you ask about the
weather
b. Often explicit, detailed questions are exactly what you want to avoid
c. Different ways of structuring propositions and discourses
i. Different cultural models
d. Good approach: get them talking
i. Ask them to talk about it or describe it to you
e. Bad approach: can we do this?
i. Can you say "Strong am I in the Force.”?
ii. Answer: yes
f. Grasp of English
i. Internationalize your English don’t use embedded structures, and
avoid conditionals or garden path sentences
7. Participant structures
a. Philips, Susan U. Participant Structures and Communicative Competence:
Warm Springs Children in Community and Classroom.
8. Conflict resolution
a. Your role in conflict should be as a facilitator
i. Encourage the language experts to interact
ii. Let them make the decision, rather than you
3
b. Who are you accountable to? What are the long-term effects of your decisions
on the linguistic success of the community?
c. Inform and educate when needed, but give the language experts the tools to
make the decisions themselves.
d. Getting to Yes by Fisher & Ury
9. Checking your data
a. ALWAYS DOUBLE CHECK YOUR ELICITED RESPONSE
i. Have them double check your notes, or read the sentence back to you
ii. "How's that sound?" elicits an acceptability judgement
1. Listen for hesitations
b. Restrict the speaker’s output as little as possible
i. This will lead to more natural language
ii. Always avoid leading questions
iii. “What are some other ways to say/talk about this?”
1. Generates a lot of ideas
2. Often gets you the data you’re looking for immediately
10. Linguistics
a. Languages tend to have linguistic terminology specifically for that language
i. Look these terms up whenever possible, and match them to broader
linguistic definitions and typologies
b. Avoid imposing (or letting your informant impose) linguistic terminology,
even if they have a linguistic background
i. Let’s face it – sometimes previous work on the language is just plain
wrong
c. Check your theories about the language with the speaker
i. Educate them a little about the linguistics behind it while you’re at it
d. Let the speaker know precisely what you’re looking for
i. Sometimes they’ll know exactly what you mean and start giving
examples
ii. Most useful when you share a common terminology
e. Beware of priming effects don't continually present the speaker with the
same types of sentences
f. Teach the informant a little about linguistics
i. Try to get the informant to think twice before making linguistic
generalizations; often they're too quick to generalize, overlooking
certain exceptions or alternate uses of a structure/word.
11. Working with teachers
a. Often language informants are former or current teachers or people who
pass on education in the community
b. Ask them “how do you typically teach this?”
i. Gets them explaining the language to you the way they explain it to
their students
4
12. Mock conversation
a. Ask them to have a conversation with each other
b. Then ask them to break it down for you
c. You begin to learn important social features of communication, e.g. how to
greet people, how to ‘small talk’, what types of questions are appropriate to
ask, etc.
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.