# The Branches and What They Are For

Languages I · Introduction to Linguistics · https://tryals.app/learn/languages-i/the-branches-and-what-they-are-for

## Linguistics Where It Touches Something Else

Linguistics is interdisciplinary by nature. Each framing of language creates its own subfield:

| Subfield | The question it adds | Typical method |
|---|---|---|
| **Psycholinguistics** | How is language processed in real time? | Reaction times, eye-tracking, priming |
| **Neurolinguistics** | Where and how is it implemented? | Lesion studies, imaging, stimulation |
| **Sociolinguistics** | How does it vary with the social world? | Interviews, sampled communities |
| **Linguistic anthropology** | What does it do inside a culture? | Long-term fieldwork |
| **Historical linguistics** | How did it get this way? | Comparison of related languages |
| **Computational linguistics** | Can it be modelled and processed? | Corpora, formal models, machine learning |

The distinction that keeps them apart is **evidence**, not subject matter. All six subfields can study the exact same word.

### Orderly variation

**Sociolinguistics** shows that language variation is not random noise. William Labov showed that pronunciation shifts with social class and conversational care. Linguistic variation is thoroughly **structured** across communities.

### The scale of the object, and its loss

There are roughly **7,000** living languages worldwide. 

Their distribution across humanity is heavily skewed:
* The largest languages account for half the world's population.
* Meanwhile, **half of all languages have fewer than 10,000 speakers**.

Many of these smaller languages are losing younger learners. This crisis makes **language documentation** an urgent scientific subfield. When a language vanishes without a record, its unique grammar is lost forever.

### Applied linguistics

Linguistic theory transforms practical fields:
* Speech technology and machine translation
* Language teaching and assessment
* Forensic authorship analysis
* Clinical treatment for speech disorders

> **Common pitfall:** treating subfields as separate topics rather than distinct methods. A psycholinguist and a sociolinguist can easily study the same dialect.

## Practice questions

8 of this lesson's 13 practice questions, with answers. The full set is in the app.

### 1. Why does Labov's department-store study show that variation is structured rather than random?

A. Higher-status sales clerks avoided regional accents entirely across every interaction
B. Rhotic pronunciation was steadily disappearing across all three historical generations
C. The same speakers shifted predictably with store prestige and with careful repetition
D. The pronunciation remained entirely uniform across all speakers within a given store

**Answer:** C. The same speakers shifted predictably with store prestige and with careful repetition

**Why:** Because the variation was **predictable in two directions at once**: across speakers by the prestige of the store, and within a speaker by how much attention they were paying. Random variation does not move when you change one variable and hold the words constant.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/why-does-labovs-department-store-study-show-that-variation-is

### 2. About half of the world's roughly 7,000 languages have fewer than 10,000 speakers.

**Answer:** True

**Why:** **True**, and the two halves of that sentence are the same fact seen twice: a small number of very large languages leaves a very long tail. The consequence is that most of the world's linguistic diversity sits in languages most people have never heard of.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/about-half-of-the-worlds-roughly-7-000-languages-have-fewer-than

### 3. Which are legitimate reasons for documenting an endangered language?

A. The community may want records for its own transmission
B. A grammar lost is a data point lost for claims about what languages can do
C. A language with few speakers is simpler and quicker to describe
D. Documentation preserves knowledge encoded in the language's vocabulary

**Answer:** A. The community may want records for its own transmission; B. A grammar lost is a data point lost for claims about what languages can do; D. Documentation preserves knowledge encoded in the language's vocabulary

**Why:** The fourth is simply untrue: some of the most morphologically intricate languages ever described have a few thousand speakers. It is worth naming because the assumption that small means simple has a long and unhelpful history in the field.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/which-are-legitimate-reasons-for-documenting-an-endangered-language

### 4. Sort each activity by whether it is primarily DESCRIPTIVE research or an APPLICATION of linguistics.

**Answer:**

- Descriptive research: Comparing verb systems across a language family, Measuring how quickly listeners resolve an ambiguous word
- Application: Building a speech recogniser for a call centre, Attributing a disputed statement to an author, Designing a reading test for eight-year-olds

**Why:** The boundary is porous in the useful direction: applications generate data that changes the description. Speech recognition is the clearest case, the failure of hand-written phonological rules to scale was itself a finding about how much of pronunciation is statistical.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/sort-each-activity-by-whether-it-is-primarily-descriptive-research-or

### 5. Approximately how many living languages are there, to the nearest thousand?

**Answer:** 7000 (within ±500)

**Why:** **About 7,000.** The count depends on where a dialect ends and a language begins, which is a political question as much as a linguistic one. Chinese and Arabic are each counted as one language in some tallies and as a dozen mutually unintelligible ones in others.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/approximately-how-many-living-languages-are-there-to-the-nearest

### 6. Arrange the stages of a language documentation project in order.

**Answer:**

1. Establish a working relationship and consent with the community
2. Record a broad range of natural speech and texts
3. Transcribe and annotate the recordings
4. Analyse the grammar from the annotated corpus
5. Deposit an archived, accessible record

**Why:** The first and last steps are the ones a beginner skips. Documentation whose recordings sit unarchived on a dead hard drive has produced a thesis and preserved nothing, which is why archiving is now a funding condition rather than an afterthought.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/arrange-the-stages-of-a-language-documentation-project-in-order

### 7. Machine translation and speech recognition improved mainly by encoding more hand-written grammatical rules.

**Answer:** False

**Why:** **False**, and the reversal is instructive. Decades of hand-written rules were overtaken by statistical and then neural models trained on very large corpora. That says something real about language, a great deal of what speakers know is distributional rather than rule-shaped, and it is one reason corpus evidence gained standing in the wider field.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/machine-translation-and-speech-recognition-improved-mainly-by

### 8. Linguistic subfields are distinguished by evidence rather than subject matter. What follows from this principle when two researchers study the same dialect?

A. Their empirical conclusions will inevitably contradict each other
B. They must examine entirely different words within the dialect
C. One conducts purely theoretical work while the other applies it
D. They ask different questions and gather distinct data types

**Answer:** D. They ask different questions and gather distinct data types

**Why:** Subfields do not divide language into isolated territories; they interrogate the same phenomena using different types of data. Overlapping subjects do not force researchers into applied roles, nor does differing evidence imply contradictory findings about the system.

Page: https://tryals.app/practice/languages-i/the-branches-and-what-they-are-for/linguistic-subfields-are-distinguished-by-evidence-rather-than