Putting It to Work

Sociolinguistics: Language, Dialect and Society

Course 1 · Chapter 10 · Putting It to Work

The first nine chapters described what sociolinguists have found. This one is about how to do it yourself, and how to read other people's claims critically: designing a small study, spotting its weak points, and deciding what you may ethically record.

The idea in a box

Sociolinguistic data has a built-in problem. William Labov's goal was to learn "how people talk when they are not being systematically observed", yet that data can only be gathered through systematic observation. This is the observer's paradox.

The Toolkit

Wikipedia's Sociolinguistics article calls the sociolinguistic interview "the foundational method of collecting data". The researcher's main aim is casual speech, "complicated by the observer's paradox". The article names these other methods:

Method	What the page says
Interview	techniques to reduce formality include asking for narratives of memorable events and interviewing subjects in groups; researchers then compare styles and correlate demographic features with speech
Participant observation, ethnography	ethnographers such as Dell Hymes and his students "often live amongst the people they are studying"
Recordings of real encounters	natural speech analysed "in detail" (conversation analysis)
Matched-guise tests	named as a method; the page does not describe how they work
Dialect surveys, existing corpora	dialectologists survey people through interviews and compile maps; existing corpora can be analysed

The Observer's paradox article adds the problem in practice: a fieldworker observing a speaker's everyday vernacular in an interview may find the speaker shifts to a more formal register, because the speaker knows the speech will be used for research. The data then fails to reflect normal speech, and the speaker would talk differently if the researcher were absent. (Chapter 4's style-shifting is exactly why this matters.)

Design a Small Study

The checklist below is my own way of turning the course's ideas into a plan; the sources above list the methods but do not give this checklist.

Pick a variable. One feature with two or more forms, such as the ING ending or the T/D deletion from Chapter 1.
Pick the social factor. Class, age, gender, style, or audience (Chapters 3, 4 and 5).
Pick a method and name its weak point. An interview risks formal speech; an existing corpus was not collected for your question.
Decide how you will count. Count how often each form appears in each style, and compare. The Sociolinguistics article describes Labov's variationist work as focused on "the quantitative analysis of variation and change within languages".
Say what would change your mind. What result would show your guess is wrong?

Worked outline (invented, for practice): Question: does a speaker use more of one form of the ING ending when reading aloud than in a casual story? Method: record the same people reading a list and telling a memorable event, then compare counts. Weak point: people who know they are recorded may be more careful in both tasks, so the gap could be understated. Would change my mind: no difference between the two tasks.

Reading someone else's study

How was the speech collected, and could the observer's paradox have changed it?
Does the claim describe a group, or is it being applied to an individual? (The Linguistic discrimination article, in Chapter 8, reports findings about perceived intelligence; a perception is not a measurement of ability.)
Is a strong word ("only", "no evidence", "success") the source's own judgment? (Chapters 6 and 9 flagged examples.)
Is the figure an estimate with a wide range (Chapter 9's 50% to 90%, Chapter 8's Esperanto counts)?

Ethics of Recording Speakers

Wikipedia's Observer's paradox and Sociolinguistics pages say nothing about ethics, so this section rests on the Informed consent article, which is general to research (clinical examples dominate it), not specific to linguistics.

Idea	What the page says
Informed consent	"a person must have sufficient information and understanding before making decisions about accepting risk"
Three components	disclosure (the information needed and checking the person understands it), capacity (able to understand it and reason about consequences), voluntariness (free from coercion, manipulation or undue influence)
When consent is waived	social science research sometimes waives consent when knowing about the study would change the behaviour being studied, subject to ethics committee or IRB review
Children	generally lack competence for fully informed consent, so parents or guardians typically give "informed permission" and minors may be asked for assent; rules differ by jurisdiction (the "mature minor" or Gillick standard in England, Australia and Canada)
Background	the Nuremberg Code "set a base international standard in 1947"

Watch out: this is a principle, not legal advice

The observer's paradox creates pressure to record people without telling them, and the page notes that consent can be waived in social science under committee review. That is a decision for an ethics committee, not for you alone. The Informed consent page says little about confidentiality or withdrawal in research (only that dynamic consent makes withdrawing easier), so I give no rules on those; for any real project, follow your institution's ethics process and the law where you are.

Course Recap

Chapter	Core idea
1 Language in Society	sociolinguistics studies how language varies with people and situations; variables and methods
2 Dialect and Accent	dialect, accent and idiolect; the standard; language versus dialect; dialect continua
3 Who You Are	sociolect, class and gender; the four models of gender and speech; AAVE
4 Style and Register	register, style-shifting, audience design and accommodation
5 Prestige and Change	overt and covert prestige, hypercorrection, the history of rhoticity
6 Many Languages in One Mind	multilingualism, code-switching, diglossia
7 Contact Languages	pidgins, creoles, mixed languages
8 Standards and Attitudes	how standards are made, linguistic discrimination, language planning
9 Endangerment and Revival	language shift, endangered languages, revitalization

One thread runs through all of it: variation is patterned, and the patterns follow social facts such as who is speaking, to whom, and with what prestige. That sentence is my summary of the course, not a quotation.

Note on sources

This chapter uses Wikipedia's Sociolinguistics (read to 100,000 of 110,102 characters, so the last part, which the fetch tool reported as navigation boxes, was not read), Observer's paradox (a short stub, read in full) and Informed consent (read to 100,000 of 138,210; it was cut off partway through the references) articles. The Observer's paradox page does not mention the "danger of death" question, group sessions or rapid anonymous observation, and has no ethics section; the Sociolinguistics page offers group interviews and narratives of memorable events instead. The Informed consent page's coverage of the Declaration of Helsinki and the Belmont Report is only a link, so I do not describe them. The study checklist, the worked outline and the recap thread are mine.

Try It Yourself

Exercise 1

Using the checklist, outline a small study of your own: a variable, a social factor, a method, its weak point, and what result would change your mind. Keep it to five short lines.

📄 View solution

Exercise 2

You want to record teenagers at a school without telling them, so their speech is natural. Using the Informed consent page, say what is wrong with the idea, what the page says about waiving consent, and who decides.

📄 View solution

Exercise 3

A news story says "a study proves people from region X speak badly". List four questions from this chapter's critical-reading list that you would ask before believing it.

📄 View solution

Chapter 10 Quick Reference

Observer's paradox: the goal is unobserved speech, but data needs observation
Methods: sociolinguistic interview (foundational), participant observation, recordings, matched-guise tests, dialect surveys, corpora
Design: variable, social factor, method and weak point, counting, what would change your mind
Consent: disclosure, capacity, voluntariness; waiver in social science needs ethics review

Course Complete

You have finished Sociolinguistics: Language, Dialect and Society. If you want to carry on, Writing Systems of the World is a planned course that has not yet been outlined, so it has no chapters yet.