Structuring a talk or paper
The broad goal of your talk or paper is to convey to your audience
your novel contribution to an existing literature. The following
format often works quite well toward that end. One reason it works
well is that it subordinates the data you have to the question you're
asking, and to prior work already done on that question. Consider
this a convenient, ready-made, and thoroughly road-tested template.
- The general question. In this section, state the general
overall question that you have chosen to pursue, and explain briefly
why this question is important. For a talk, this can be as short as 2
or 3 sentences, but they're important ones. E.g. "My research asks
whether grammatical representations are affected by linguistic usage.
This question has attracted attention recently, because it opposes two
very different views of the nature of grammar."
- Literature review. What existing theories attempt to answer
the question raised in Section 1? Briefly describe the major
theoretical positions in the literature, and existing evidence that
supports or challenges them. Again, for a short talk, this can be
brief but it must be there.
- Open specific question. In what sense is the debate over
the general question from Section 1 not yet resolved - i.e. what
important specific question is left open by the literature you have
just reviewed in Section 2?
- Research. Present your research, or research plan, for
answering the open specific question from Section 3. Please be
concrete about methods used. Before showing us any data, describe
what specific pattern of possible results would answer the open
question from Section 3, and how. Then you can present your actual
results (if you have them), which will provide an answer to the
question.
- Discussion and conclusions. Specify what one may conclude,
given the outcomes from your research (or, if proposed research, from
different possible outcomes). Make sure you explain the broader
meaning of the possible results and how they support or challenge
existing theories.
There are some sorts of research where this template may not apply
readily - for instance, descriptive work on an endangered language,
where the point is not so much to contribute to an ongoing theoretical
dispute, but instead to capture a language before it vanishes forever.
In such cases, the "literature review" section can be given over to
other sorts of relevant background, e.g. where the language is spoken,
how many speakers, why it is of particular importance. But even in
such cases, do try to frame the talk in terms of broad goals
acccessible to everyone: "Analyzing endangered languages is important
because...", then gradually narrow in to the specifics, and finally
broaden back out to end on a suitably general note.