Lost in Translation

Yesterday, Reuters reported on the UN biodiversity conference COP17 concluding in Armenia’s capital Yerevan.¹ Nearly 200 countries agreed to accelerate efforts to halt the loss of nature by 2030. COP17 itself was the first review of the international targets adopted in Montreal in 2022,² and it found that nearly 90 per cent of countries had set individual national targets, but fewer than half of those assessed were on track – and the money remains short.

»This isn’t a data problem. It’s a translation problem«

The headline ambition is known as »30 by 30« – conserve at least 30 per cent of the planet’s land, inland waters, and seas by the year 2030.² A second target sets a matching number for ecological repair: restore 30 per cent of degraded ecosystems by the same year.² But what does it mean to conserve 30 per cent of biodiversity? And how would we know whether we had succeeded?

The problem begins with the words themselves, starting with »nature«. That word suggests a world separate from human activity, even though humans and everything we build are part of the same living world. That distinction is useful, perhaps, but it’s still a distinction that we impose. »Biodiversity« sounds more precise, yet it describes three different things. Genetic diversity is the variation within a species that allows populations to adapt to change. Species diversity is the variety and relative abundance of species. Ecosystem diversity is the variety of habitats and the relationships between the organisms and environments within them. These dimensions are connected, but not interchangeable. A forest can gain species while losing genetic diversity and the specialists that make it distinctive. Species counts alone tell us little about its resilience or ability to survive change.

The 30 by 30 slogan measures something that can be counted: area. But a new plantation and an ancient forest can occupy identical hectares while supporting radically different life. The framework’s own text adds qualifiers – »ecologically representative«, »well-connected«, »equitably governed«² – but in practice, it comes down to a percentage of territory.

Consider Chernobyl. Since the nuclear disaster in 1986, the exclusion zone has drawn back wolves, wild horses, and other large mammals. A 2015 winter track census found wolf density seven times higher than in surrounding nature reserves.³ Yet research by the evolutionary biologists Anders Møller and Timothy Mousseau found insect and spider abundance declining with radiation,⁴ and a later meta-analysis of 45 studies found elevated mutation rates across 30 species.⁵ Both can be true at once: wildlife can be abundant while radiation still damages particular organisms. What looks like recovery depends on what you choose to count.

This isn’t a data problem. It’s a translation problem.

Sociologists Wendy Espeland and Mitchell Stevens call it commensuration: converting different qualities into one comparable metric.⁶ That lets governments negotiate over nature without agreeing on every ecological detail – but once biodiversity is translated into hectares, the measurement takes on a life of its own. It becomes possible to report progress on protecting nature without demonstrating that nature is in fact recovering.

There is also a parallel problem – enforcement. No global authority can fine a country for failing to restore its forests. So what makes an agreement like this effective? Abram and Antonia Chayes call their answer the »managerial model« of treaty compliance: agreements work less through punishment than through repeated reporting and public scrutiny.⁷ Governments are expected to report their progress, explain their failures, and account for the gap between promises and results. Legal scholar Cass Sunstein’s idea of the expressive function of law offers another answer: a treaty declares what a country considers acceptable, and what it wants people to believe it stands for.⁸ The COP17 review puts both mechanisms to work. It makes governments’ promises and their progress, or lack of progress, public.

It would be easy to conclude that the targets are meaningless, or the entire process futile, because nothing is enforced. But that doesn’t follow. Without a shared measure, countries could hardly coordinate at all. Without regular reporting, they would have less reason to account for their failures. An imperfect measure can still be useful, provided we remember what it leaves out.

The danger isn’t that we count hectares. It’s believing that hectares are what we set out to save. Biodiversity isn’t a quantity waiting to be divided into percentages. It’s a web of relationships and dependencies, many of which we have yet to discover. We need numbers to make collective commitments possible. But measurement should help us understand what we are protecting, not reduce protection to something we can count.

References