As teachers a large portion of our time is spent on rules. Helping learners uncover the rules that underly English, challenging the rules they’ve overgeneralised, & occasionally admitting that a rule isn’t really a rule at all. But step back from the classroom for a moment & a bigger question appears: how many of the rules we teach are truly universal, & how many are just English being English?
A recent paper in Nature Human Behaviour by Annemarie Verkerk, Russell Gray & colleagues at the Max Planck Institute for Evolutionary Anthropology tackles that question head-on, using the world’s largest grammatical database to test which “universal” patterns actually hold across 1,700 languages.
The study
The team analysed 191 proposed linguistic universals using Grambank, a typological database covering 400+ grammatical features across the world’s languages. Crucially, they didn’t treat languages as isolated datapoints. Instead, they used Bayesian spatiophylogenetic modelling to control for ancestry (shared historical roots) & geography (contact-driven influence). This matters: English & Dutch share features not because those features are universal, but because they share a history.
Each language was coded for features such as word order, case marking, agreement, negation strategies & clause structure. The researchers then tested whether each proposed universal still held once genealogical & geographical biases were removed.
The findings
Around one third of the proposed universals showed strong statistical support. Many “rules” we assume are cross-linguistic tendencies turn out to be family traits or regional patterns rather than human universals.
A few examples:
- Languages with flexible word order often rely on case marking to keep meaning clear, while languages with fixed word order tend to use less case.
- If a language marks plural on nouns, it’s more likely to mark number on pronouns too.
- Languages with obligatory subject expression tend to have clearer verb agreement systems.
So what does this mean exactly? Think of Finnish: rich case marking, flexible word order, clear agreement. Compare that with Vietnamese: minimal case marking, relatively fixed word order, no verb agreement. These patterns aren’t coincidences; they’re structural tendencies that recur across unrelated languages.
English itself is a good reminder: fixed word order, minimal case marking & limited agreement make it typologically unusual compared with many of the world’s languages.
Context
This work sits in a long tradition stretching from Joseph Greenberg’s typological universals to contemporary probabilistic approaches in linguistic evolution. It also resonates with usage-based linguistics (Bybee, Tomasello), which argues that grammar emerges from pressures of communication, processing & interaction rather than from fixed innate templates.
Teacher takeaways?
- When learners struggle with English word order, remember that their L1 may encode relationships differently (through case, agreement or discourse).
- “Universal grammar” is less universal than once thought; learners bring structurally diverse expectations to the classroom.
- Highlighting cross-linguistic patterns (e.g. how languages signal who does what to whom) can help learners map English structures onto familiar conceptual territory.
What fascinates me here is how deeply languages vary yet how consistently certain pressures shape them. It’s a reminder that grammar is not arbitrary: it’s an adaptive system shaped by history, contact & human cognition.
What grammatical patterns from other languages have surprised you?
Leave a Reply