Courses:

This course is designed to allow participants to engage in the exploration of the grammatical structure of a language that is unknown to them (and typically to the instructors as well). In some ways it simulates traditional field methods research. In terms of format, we work in both group and individual meetings with the consultant (students are encouraged to work in teams of two to three, to minimize the number of individual meetings for the consultant). Each student identifies some grammatical construction (e.g. wh-questions, agreement, palatalization, interrogative intonation) to focus their research: they elicit and share data and write a report on the material gathered that is to be turned in at the end of the term. We can then put together a volume of papers exploring various aspects of the grammar of the language. The first three to four weeks of the term, our group meetings explore the basic phonology, morphology and surface syntax for a first pass overview of the language, looking for interesting areas to be explored in more detail later. During this period individual sessions can review material from the general session as well as explore new areas. At roughly the fifth meeting, individual students guide the group elicitations to explore their research topic.

The course does not use a textbook, and direct instruction in elicitation techniques during class time is kept to a minimum; students learn by watching the instructors, and by doing elicitation themselves under the instructors' guidance. Some more explicit instruction appears in the form of a running commentary on the class website; this has the advantage of not taking up class time with methodological discussion, and also of making it easier to plan elicitation strategies without the consultant's knowledge.

Techniques which are typically discussed include

• Breaking complex questions down into smaller, more simple structures
• Making judgments easier by asking for a comparison of two minimally different examples (rather than simply asking whether a sentence is well-formed in isolation)
• Using pragmatic bias to sharpen judgments about potentially ambiguous sentences
• Finding out why crucial sentences have been rejected (that is, bearing in mind that a sentence may have been rejected for reasons unrelated to the phenomenon the investigator is interested in; consultants very often do not distinguish, unless asked, between syntactic, semantic, pragmatic, and phonological ill-formedness)
• Avoiding leading questions
• Avoiding, whenever possible, open-ended questions (like "in what context would you say X?"), in favor of questions set in a particular context (like "can you say X in context Y? or context Z?")
• Making the consultant a full participant in the process by explaining (to the extent that the consultant is interested) what questions are being addressed, and taking the consultant's theories about these questions seriously (without necessarily accepting them as the last word)