“The rise and fall of language diversity through the Holocene” by Blasi, Hamilton, Gray & Bowern, published in Science last month, does something I love. It complicates a standard narrative.
The received story of language diversity is that it has been declining steadily for centuries, largely driven by colonialism and globalization. But direct evidence is scarce: writing only reliably takes us back about 6,000 years. So the team combined ethnographic data from 171 hunter-gatherer and fisher societies, paleodemographic estimates of global population, and statistical and social-computational modelling to reconstruct plausible trajectories of global language diversity over the 12,000-year Holocene.
At the beginning of this period, they estimate there were roughly 4,500–6,200 languages. That’s actually fewer than the roughly 7,500 languages spoken today. Language diversity then increased alongside population through most of the Holocene, reaching a peak between about 1,000 and 3,000 years ago, when the models suggest that tens of thousands of languages may have existed — perhaps an order of magnitude more than today.
Then came a major collapse. The authors argue that the decline in linguistic diversity began with the expansion of large states and empires, long before European colonialism. Expanding populations and their languages, cultures and institutions displaced smaller ones. Today’s languages are therefore “survivors of a massive and highly selective historical bottleneck.” They persist because the populations speaking them expanded. Extinction, the authors argue, has shaped linguistic and cultural diversity far more than we have appreciated.
Now, there are obvious caveats. The authors are modelling linguistic diversity, not measuring it. Their estimates depend on heroic assumptions about early ethnolinguistic group sizes, global population and the relationship between population and language number. Regional histories could have been very different, and the model can’t reconstruct individual episodes of extinction caused by war, disease or environmental change.
But the paper provides a provocative template for thinking about diversity more generally. It suggests that population growth can initially generate diversity; expanding societies can subsequently erase it through homogenization.
And that makes me wonder about agricultural biodiversity. You saw that coming, I’m sure.
I’m not suggesting that crop diversity followed the same pattern as language diversity through the Holocene. We simply don’t have the evidence to say that, though it’s an intriguing thought. But could we borrow the methodology? Could we reconstruct plausible trajectories of crop or agricultural-system diversity from fragmentary archaeological, ethnobotanical, genetic and demographic evidence?
Rather than assuming that agricultural diversity was vast and stable in the distant past in its Vavilovian centres, and has been declining recently, perhaps we might end up complicating the familiar genetic erosion narrative too. That would be worthwhile. And fun.