With plans for REF 2020 well under way in institutions across the UK, there’s a great deal of debate about what makes the best ‘research output’ and how best to understand, and secure, ‘research impact’. But I’m struck by how little contestation there is over the term ‘research environment’ – and yet doesn’t this term speak directly to the landscapes upon which research takes place: its cultures, contexts and conditions of possibility? An environment can be dead and dusty and barren; it conjures up images of the future sequences in Terminator: full of machines. But what if we though about ‘research ecologies’ instead? Ecologies are living and breathing and vibrant and diverse. They are organic and unpredictable and they change over time. Research involves creativity, in all of that sense (quite literally the creation of something); I imagine colourful gardens – or jungles – full of birdsong and monkey squeals and smelly plants. And I think about how, with a garden (or jungle), you cannot poke and prod at the plants and say ‘You absolutely must grow more’. Ecologies have to be left to breathe – there’s a sense of delicate balance, and of the need to give space, and to nourish. And, conversely, ecologies are very easily destroyed (not least by poking and prodding, but also by not enough light, or water, or air) … Although perhaps the analogy is a bad one, given that we’re systematically destroying ecological systems. Still, I can’t help but imagine it: what if we thought in terms of research ecologies instead?