# Universals and Typology

Languages I · Introduction to Linguistics · https://tryals.app/learn/languages-i/universals-and-typology

## Unity Underneath Diversity

Languages differ immensely, but not at random.

* **Typology** classifies languages by their structural properties.
* The search for **universals** asks what constrains human language variation.

Linguists distinguish two main approaches:

* **Taxonomic (Greenbergian) universals**: empirical generalisations induced from a sample of attested languages.
* **Theoretical (Chomskyan) universals**: innate principles attributed directly to the human language faculty.

### Greenberg's Method

In 1963, Joseph Greenberg sampled 30 languages and proposed 45 universals. The most powerful statements were **implicational**:

> If a language has property X, it also has property Y.

These statements forbid one logical combination. A single counter-example directly refutes the claim.

### Basic Word Order

All six word orders exist, but frequencies vary wildly across languages:

| Order | Share |
|---|---|
| SOV | ~41% |
| SVO | ~35% |
| VSO | ~7% |
| VOS | ~2% |
| OVS | <1% |
| OSV | <1% |
| None | ~14% |

**Subject-before-object** order dominates, appearing in roughly 86% of ordered languages. 

Word order also **correlates** across elements. For example, VO languages typically use prepositions, while OV languages prefer postpositions.

### Absolute vs Statistical

* An **absolute** universal admits zero exceptions across all human speech.
* A **statistical** universal represents a strong, widespread cross-linguistic tendency.

> **Common pitfall:** Confusing the sample with the population. A biased language sample can easily disguise statistical tendencies as universal rules.

## Practice questions

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

### 1. Approximately what percentage of languages with a dominant word order place the subject before the object?

**Answer:** 86 (within ±6)

**Why:** **About 86%.** SOV and SVO between them take roughly three quarters of all languages, and adding the smaller subject-first orders brings the total to around 86% of those with a dominant order. The verb's position, by contrast, varies freely, the constraint seems to be on the arguments.

Page: https://tryals.app/practice/languages-i/universals-and-typology/approximately-what-percentage-of-languages-with-a-dominant-word-order

### 2. All six logically possible orders of subject, object and verb are attested in some language.

**Answer:** True

**Why:** **True**, though the last two are vanishingly rare, a handful of languages each, several of them Amazonian. Their existence matters out of proportion to their number: it turns "subject before object" from an absolute universal into a statistical one, which is a different kind of claim.

Page: https://tryals.app/practice/languages-i/universals-and-typology/all-six-logically-possible-orders-of-subject-object-and-verb-are

### 3. Sort each property by whether it is typical of VERB-OBJECT languages or of OBJECT-VERB languages.

**Answer:**

- Verb–object type: Prepositions, Noun before relative clause, Auxiliary before main verb
- Object–verb type: Postpositions, Relative clause before noun, Main verb before auxiliary

**Why:** The pattern is the harmony: whatever a language does with verb and object, it tends to do with adposition and noun, noun and relative clause, auxiliary and verb. Whether this reflects one parameter, parsing efficiency or historical inheritance is unresolved, and the tendencies are strong rather than exceptionless. English itself puts relative clauses after the noun and adjectives before it.

Page: https://tryals.app/practice/languages-i/universals-and-typology/sort-each-property-by-whether-it-is-typical-of-verb-object-languages

### 4. Why might a universal induced from a sample of languages be less secure than it looks? Select every legitimate concern.

A. The sample over-represents well-described and written languages
B. A sample of 30 languages is too small to state anything
C. Areal contact can spread a feature across unrelated neighbours
D. Genetically related languages may share a feature by inheritance rather than by constraint

**Answer:** A. The sample over-represents well-described and written languages; C. Areal contact can spread a feature across unrelated neighbours; D. Genetically related languages may share a feature by inheritance rather than by constraint

**Why:** The first three are exactly why typological samples are stratified by family and region. The fourth overstates: Greenberg's 30 were chosen for genetic and areal spread rather than convenience, and the implicational results have largely survived samples fifty times larger.

Page: https://tryals.app/practice/languages-i/universals-and-typology/why-might-a-universal-induced-from-a-sample-of-languages-be-less

### 5. Complete the account of typology.

**Answer:** A universal induced from a sample of attested languages is a **taxonomic** universal, while one attributed to the language faculty is a **theoretical** one. A statement of the form "if a language has X it also has Y" is **implicational**. The most common basic word order is **SOV**.

**Why:** SOV at about 41 per cent is the answer students most often get wrong, because English is SVO and the languages most often taught in Europe are too. Japanese, Turkish, Korean and Hindi are all verb-final, and the type is the most common on earth.

Page: https://tryals.app/practice/languages-i/universals-and-typology/complete-the-account-of-typology

### 6. Greenberg's 1963 study examined thirty languages selected for genetic and areal spread.

**Answer:** True

**Why:** **True**, and the deliberate selection is what made a sample that size worth anything. Thirty languages from one family would show inheritance; thirty from one region would show contact. Spread is what lets a recurring pattern be attributed to something about language rather than to history.

Page: https://tryals.app/practice/languages-i/universals-and-typology/greenbergs-1963-study-examined-thirty-languages-selected-for-genetic

### 7. Greenbergian universals are induced from empirical samples, whereas Chomskyan universals are posited as innate mental constraints. What does this distinction imply when a newly documented language violates an absolute universal?

A. Both theoretical frameworks must discard the underlying principle as wholly invalid
B. A Chomskyan claim is falsified whilst a Greenbergian tendency accommodates the data
C. A Greenbergian rule is immediately refuted whilst a Chomskyan principle can survive
D. Neither framework is affected because statistical exceptions are routinely tolerated

**Answer:** C. A Greenbergian rule is immediately refuted whilst a Chomskyan principle can survive

**Why:** Because taxonomic universals are empirical claims across attested speech, one exception disproves an absolute rule. Conversely, innate generative constraints may remain theoretically viable by attributing surface violations to language-specific parameters or performance factors rather than immediate refutation.

Page: https://tryals.app/practice/languages-i/universals-and-typology/greenbergian-universals-are-induced-from-empirical-samples-whereas

### 8. What percentage of the world's languages have SOV as their basic word order?

**Answer:** 41 (within ±8)

**Why:** About **41%**: the largest single group, with SVO close behind at around 35%. Neither is a majority, which is worth noticing: the most common word order on earth is used by fewer than half of all languages, and the popular assumption that SVO is the norm reflects which languages Europeans encounter rather than any global fact.

Page: https://tryals.app/practice/languages-i/universals-and-typology/what-percentage-of-the-worlds-languages-have-sov-as-their-basic-word