Naomi Nagy

Linguistics at U of T


Writing sociolinguistics papers

Here are two sets of guidelines (one short and one long) for writing a sociolinguistics paper.

In brief:

  1. Present your hypothesis or hypotheses.
  2. Be sure to write these in a "testable" manner-- no hedging.

  3. Describe your methodology.
  4. Be detailed enough that someone could replicate the experiment/study. (This is necessary to see how robust your findings are.)

  5. Present the data in a clear chart, list, or table (maybe more than one).
  6. This should correspond to your hypothesis.

  7. Discuss what you found.

Does it support your hypothesis?

If not, why not?

What might you do differently if you were to continue this research project?

In more detail:


What issue(s) is/are at stake?

What questions will be addressed?

What is your hypothesis? Why does that matter?

What type of data will be used?

Lexical, phonological, syntactic... ?

Spoken, written, heard (like the Gating experiment)...?

What is/are  the variable(s)? What are the possible variants of each variable?

Describe the speech community to be examined in your project.

Who is part of the community? Who isn't? That is, how will you decide who can/can't be an informant for this project? Do they all share some social factors?

Why do they form a cohesive group worth studying?

What can be learned by studying this community?

How are you qualified to be the investigator?

What type of analysis will be conducted?  

Briefly, what conclusions are drawn?


Who are the informants? Who are the investigators? How many of each are there?

Explain the method you used to locate possible informants and get them to participate.

What did you tell them? What was their reaction? (Did your method work well?)

Was the data recorded, transcribed, acoustically analyzed...? By what method?

How many questions (or items elicited)? Where? When?

How were the results analyzed?

Was statistical analysis involved? What type? Why? 


Dependent variable

This is the place to describe the variable(s) you are investigating.  Objectively explain what the different variants are.

Define the envelope of variation for the dependent linguistic variable(s).

Avoid counting tokens that, due to some aspect of their linguistic, social, or stylistic context, do not allow for variation. For example, if you are studying the (ING) variable, you need to avoid counting monomorphemic words such as "ring," since no one ever says [rIn]. Explain all such decisions you make.

Independent variables

There are several types of variables that may affect/be correlated to the dependent variable you are investigating.  These may be grouped into three sets: social (e.g., age, gender, ethnicity), stylistic (e.g., casual conversation, reading a passage), and linguistic (e.g., preceding context, position in sentence, height of vowel).  

Social independent variables

Define each of the social factors that are relevant to your project.

Explain how you will select the proper code/variant each informant for each factor.

Is the criteria objective?

If not, do the informants or the investigator make the necessary decisions?

What steps did you take to make sure that you can evaluate the effect of each factor?

Provide a table showing how you coded each informant for each social factor.

Stylistic independent variables

Are there stylistic variables?

If so, how will each token be coded for style? Give some examples.

If not, are you taking steps to control for stylistic variation? Why or why not?

How many different styles will you consider? What are they? Why?

Linguistic independent variables

What linguistic factors were considered (that might influence the outcome)? phonetic, phonological, syntactic...

Define each independent linguistic variable.

Explain what the possible variants are for each.

 Describe them using IPA if appropriate, or using theoretical linguistic apparatus (e.g., trees or brackets for syntactic variables).

Is it clear that all the variants mean the same thing? If not, why do you consider them as one variable?

What method(s) will you use to code each variable?

What effect do/did you expect each variant to have? Why? Were you right? If not, why not?

Provide a table/list that summarizes all the relevant linguistic variables and their possible variants, with examples from your data.

Either provide examples of some coded tokens in the text of your paper, or an appendix with all of the coded data.


Analyze the data and discuss the patterns found. Earlier assignments prepared you for this part.

What did you find out? Discuss each independent variable individually, commenting on how it correlates to (each/the) dependent variable.

Was there any interaction between the variables?

Organize the section so that the relevance to your hypotheses is evident.

Discussion / Conclusion

What did you find out? What's the most important "take-home message"?

Do your data support your hypothesis?

Be sure to respond to everything you raised in the introduction.

What surprises did you find?

What did this investigation suggest for future research?

How might the study be improved?

You are strongly encouraged to turn in drafts of your papers (or some sections of your papers) at least 1 week BEFORE the due date. I will be happy to read them and give you feedback so that you can revise and resubmit. Only the last version will be graded.

NO CREDIT will be given for late papers.

Those of you who are submitting proposals for large scale projects (an option in some courses) still need to collect and analyze some data, otherwise you (and I) won't know if your methodology works. Also, even small pilot sets of data should illustrate the correlations you expect to find. If so, talk about that; if not, explain why not.


Updated April 3, 2017.

email: naomi dot nagy at utoronto dot ca | Return to my home page