I’ve heard it said that languages borrow vocabulary easily, but grammar stays put. Words travel, structures resist. This would explain the existence of loanwords & why English happily absorbed piano & ballet, but didn’t import Italian verb conjugations or French gender agreement. This idea isn’t just pedagogical folklore. It comes from decades of linguistic theory. However, when a recent study took a look at language contact globally, the narrative appeared to unravel.
The study
Graff et al. asked a bold question: when languages come into contact, how much do they actually influence each other, & is this influence predictable?
The challenge is that historical contact is hard to document. We rarely have clear records of who spoke what to whom centuries ago. So the researchers used a clever workaround: genetic admixture. If two populations show genetic mixing, it strongly suggests prolonged social contact between adults, exactly the conditions under which languages can influence one another.
They also examined language pairs, meaning two different languages whose speaker populations are known to have been in contact. Importantly, these languages were not from the same family. So the similarities they observed couldn’t be explained by shared ancestry, only by contact.
For example, instead of comparing Spanish & Italian, they compared pairs like Spanish–Quechua or Bantu–Khoisan. This allowed them to isolate contact effects rather than inheritance.
Across more than 370 languages worldwide, the researchers compared:
- Language pairs with evidence of population contact
- Random pairs of unrelated languages with no known contact
Rather than focusing on obvious loanwords, they looked at structural features: word order, tense systems, sound patterns, stress & meaning organisation. In other words, the parts of language we often assume are hardest to shift.
The Findings: A Subtle, Unpredictable Influence
At first glance, contact does what we expect: it makes languages a bit more similar. But the size & nature of that effect are surprising. On average, contact increased shared structural features by only a few percentage points. That might sound small, but across hundreds of features & languages, it’s a consistent global pattern.
More interestingly, what gets borrowed is highly uneven, & the direction of change isn’t always toward similarity. Some structural features spread across languages quite readily. Others barely move at all. In a significant number of cases, languages in contact actually become less similar in certain areas. This happens when communities deliberately emphasise difference as a marker of identity.
So the familiar hierarchy of “words first, grammar last” doesn’t hold up at a global scale. Contact doesn’t follow a universal rulebook. It follows people.
How Does Grammar Even “Borrow”?
The study documented the outcome of this structural convergence & divergence, but historical data limits exactly how we see the process. Based on wider linguistic research, however, we can understand the pathways. Grammar isn’t “borrowed” like a loanword; it shifts through sustained interaction. This can happen by:
- Calquing: Translating a grammatical construction piece-by-piece (e.g., adopting a new way to form the perfect tense).
- Borrowing function words: Adopting conjunctions or prepositions that bring new syntactic patterns.
- Solidifying an option: Making a rare word order in one language (like Subject-Verb-Object) more dominant & rigid under influence from a contact language that uses it.
- Simplification: Losing complex, redundant features (like grammatical gender) that aren’t critical for understanding in a multilingual environment.
Crucially, the study’s method reveals the net result of all these possible pressures—along with the counter-pressure of identity-based divergence. It shows that grammatical systems are permeable, but the changes are typically probabilistic, subtle, & socially negotiated.
More interestingly, what gets borrowed is highly uneven.
Some features spread across languages quite readily. Others barely move at all. Interestingly, in a significant number of cases, languages in contact actually become less similar in certain areas. This happens when communities deliberately emphasise difference as a marker of identity.
So the familiar hierarchy of “words first, grammar last” doesn’t hold up at a global scale. Contact doesn’t follow a universal rulebook. It follows people.
Why this matters for ELT
For teachers, this challenges some deeply ingrained assumptions about learner language.
If grammar really were immune to contact, we’d expect learner grammars to stabilise quickly around target norms. But in reality, learners often develop systematic, socially meaningful patterns that don’t disappear with more input.
This research aligns well with usage-based & sociocultural views of SLA: grammar is not a sealed system acquired once in childhood. It remains flexible, sensitive to frequency, social pressure & communicative need.
It also helps explain why certain “non-standard” features persist in multilingual Englishes worldwide. These aren’t simply fossilised errors. They’re outcomes of contact, adaptation & identity work.
Teacher Takeaways?
This research doesn’t offer ready-made classroom strategies. Instead, it challenges some comfortable assumptions.
- Grammar is not immune to contact or adult learning, even if it feels that way in the classroom. Large-scale evidence shows that grammatical systems can shift through sustained interaction, not just early childhood acquisition.
- Persistent “non-standard” patterns may reflect stable contact effects, not failure to progress. When such patterns are shared across learners or persist over time, they may signal an emerging norm rather than incomplete learning.
- Learner English is shaped by social alignment as much as by input & instruction. Choices about grammar, pronunciation or rhythm can function as identity signals, not simply gaps in knowledge.
As food for thought, it’s a reminder that learner language is often doing something purposeful—negotiating meaning, aligning with a peer group, or signalling identity—even when it diverges from the syllabus.
How do you see language contact shaping the English used by your learners?
Leave a Reply