Why does Australia almost never change its Constitution?
Two Swiss farmers forced a national vote about cows' horns. Australians can't ask a question at all, and haven't said yes since 1977. Here's why.

Art: Omxus Press

Cover of The Language Proof (Omxus Press)
In Paris around the turn of this century, researchers played recordings to adults born in Korea and adopted by French families as children. Some sentences were Korean: the language they'd first heard, from the people who first held them. Would anything in them recognise it?
They couldn't pick it. In tests of memory for Korean, the adoptees did no better than French people who'd never heard a word of it. The researchers put them in a brain scanner too. Their brains showed no special response to Korean over other unknown languages, and French lit up much the same areas as it did in people born to it (Pallier et al., 2003).
A language is the most demanding thing most of us ever learn: tens of thousands of words, rules nobody can state, sounds tuned by a first birthday (Kuhl, 2004). And it can come entirely from the room a child grows up in. Swap a baby from Tokyo with a baby from Perth and each speaks the other's language with no accent at all. Nobody disputes it. It's so obvious it feels trivial, and that's the point.
Genes matter for how fluent a person becomes (Stromswold, 2001). But no twin study estimates the heritability of speaking Japanese rather than English, because where the twins grew up fixes the answer. A capacity can be inherited while what fills it comes from outside.
The US census asks every member of a sampled household what language they speak at home, where they were born, when they arrived and how they are related to the householder. We used the 2022 file for California: 373,942 people aged five or over with a language answer (U.S. Census Bureau, 2023).
Among Californians of the same ancestry, those born in the US were far more likely to speak only English at home than those born abroad. Filipino ancestry: 82.6% against 13.3%. Chinese: 56.1% against 9.3%. Korean: 51.6% against 12.5%. Mexican: 35.0% against 5.2%. Across six ancestries the gaps run from 29.8 to 69.3 percentage points. The ancestry is held constant. The birthplace changes, and the language changes with it.
The sharper test is children adopted from abroad, who share no genes with the household. Ninety-six children aged 5 to 17 in the sample were born abroad and adopted by the householder, most often from China, Guatemala, Mexico, Ethiopia, Russia and Vietnam.
| Born abroad, adopted, aged 5 to 17 | Speak only English at home |
|---|---|
| Householder speaks only English (60 children) | 54 (90.0%) |
| Householder speaks another language (36 children) | 5 (13.9%) |
An adopted child's home language matched the householder's 88.5% of the time. Biological children's matched 87.1% of the time (Fisher exact p = .88). If shared genes pulled a child towards the parents' language, biological children would match more often. They do not. Sharing genes with the household added nothing we could detect.
Among 96,114 foreign-born adults, 78.6% of those who arrived aged 0 to 4 speak English only or "very well". The share is 59.6% for arrival at 10 to 14, 40.5% at 15 to 19 and 35.0% at 30 or older. A 1990 census study of 2.3 million Chinese- and Spanish-speaking immigrants found the same decline (Hakuta et al., 2003).
When a child in Sydney speaks English, nobody wonders about an English gene. We don't say the kid gets it from bad blood. Yet when the talk turns to temper, trust or crime, genes and choice become the default explanation. Courts in several countries hear genetic "predisposition" evidence. Defence experts have testified about MAOA genotypes at murder trials in the United States (Bernet et al., 2007), and an Italian appeal court cut a murder sentence partly on that evidence (Forzano et al., 2010). A review of more than 1,500 US judicial opinions found neuroscience and behavioural genetics routinely raised in criminal cases (Farahany, 2016).
If the most complex behaviour people perform follows the home and not the genes, a claim that a simpler pattern is in the genes needs its own evidence. Very little of it meets the standard the language case meets. And learned means relearnable. Those children in Paris took on a whole new language because someone took them home and talked to them.
This is observational, one US state in one year. Only 96 foreign-born adoptees appear in the sample: enough for effects this large, not for splits by birthplace. The census asks about language spoken at home, and the householder's language stands in for the parents'. Age at arrival is tangled up with education and reasons for migrating, which is why the gradient, not any single step, is the finding. We do not claim genes are irrelevant, that every behaviour is like language, or that people lack agency. We claim a default: until shown otherwise, assume a complex pattern was learned.
The first language may not vanish altogether either. A later study scanned teenagers adopted from China in infancy and found traces of Chinese tones in their brains, years after they'd lost the language (Pierce et al., 2014). Early experience leaves marks. That's rather the point: what marks us is what happened to us.
Next time you hear "it's in the blood", "the apple doesn't fall far" or "kids like that", ask one question: where would they have learned it? The answer doesn't excuse anything. It tells you where to work.
And if you've got a child anywhere near you, talk to them. Read to them, sing badly, narrate the shopping. Whatever they're hearing from you is going in, and they aren't stuck with it, or with what you heard. Nor are you. People take on a whole new way of speaking when someone makes room for them.
Two Swiss farmers forced a national vote about cows' horns. Australians can't ask a question at all, and haven't said yes since 1977. Here's why.

Art: Omxus Press
A small lottery lifted student turnout by 32%. Paying for blood drove women away. What moves us is rarely what we say it is, and somebody designs it either way.

Art: Omxus Press
A dead salmon lit up in a brain scanner because the method expected it to. When a clinician's belief decides the finding, what choice is always safe?

Cover of The Lonely Whisper of the Wind (Omxus Press)