Not long ago, I co-wrote a grant with a small army of of collaborators. As our deadline approached, we were circulating several e-mails a day among 24 people at 8 different institutions around the United Kingdom. Each 'reply-all' message laboured under the size of the latest attachment. Time stamps reflected ever-blearier hours. Our ambition for a major interdisciplinary consortium was at stake.

Our project called for researchers from the natural sciences (in hydrology, soil, ecology, sediment transport, water chemistry and gas exchange) to join colleagues from the social sciences (in economics, politics, policy and governance) to study changes in environmental resources at a national scale. We would collaborate to examine managed landscapes holistically, as integrated systems of natural processes and human activities. We would be a supergroup united by data sets, computer models, statistical analyses and case studies.

As iterations of the proposal rolled through our inboxes, it occurred to me that for all our use of future tense and conditional clauses, the promise of what we would do if funded, this — the collaborative act of writing the proposal — was really where our interdisciplinary work was happening. The requirement of writing a joint proposal was forcing us to transcend the shorthand of our respective disciplines.

The proof was in the rainbow of accumulated tracked changes and marginalia. “We're going to drive this model with results from that one, right?” someone asked in electric green. Yes. Terms familiar to one contributor were foreign to another. Clauses such as “construct a platform for linked subroutines capable of accounting for fluxes through the selected grid space” disappeared. “Deliberately abstracted” became “simplified”, and “dynamically integrated comprehensive framework” became “system”.

Co-authors haggled over content. Does a group of integrated models comprise a single model? Are we simulating or predicting or forecasting? Should we emphasize the elements that make the work applied, or the unresolved dynamics that make it exploratory? Each description, each choice of evidence, each reference was weighed. Debates about what stayed in a draft and what did not were essential to our internal process of deciding what we actually wanted to do. They propelled simpler, clearer writing — to everyone's benefit. By the time our submission was ready, what had started as a sprawling pitch had transformed into a compelling plan. The steps, however technically complicated, felt like items that we could tick off of a to-do list.

This grant proposal marks my fourteenth formal interdisciplinary undertaking. Each one has been an education. You hear the language of another discipline: its phrases and idioms; its favourite verbs and adjectives. You learn how colleagues from other backgrounds construct and solve problems. You gain a sense of the ideas that interest them, and an appreciation of why. You hear the language of your own discipline as if it were someone else's. You revisit your own preferences for puzzles and questions. You strike your foregone conclusions, hang up your implicit assumptions and begin to build — and rebuild — explanations and arguments from scratch.

Every day, I read posts from institutions around the world soliciting ways of cultivating interdisciplinary initiatives. But I have found that strategic discussions about interdisciplinary research rarely come down to the practicalities of the research process.

Interdisciplinary research is too often described in terms of lofty abstraction, a collaborative effort that sounds more magical than methodical. Writing together makes projects real. It requires negotiations between disciplines and individuals (see Careers Column page 427). The exercise closes conceptual distances that conversations leave wide open.

The next time I am invited to an interdisciplinary workshop to galvanize links between researchers from disparate specialities, I might ask the organizers how much collaborative writing is on the agenda. If the event is mine to organize, I'll scrap hours of showcase presentations. Attendees will write. I'll offer a handful of relevant prompts: calls for research proposals from a funding agency, advertisements for an interdisciplinary conference, the announcement of a journal's special issue. Participants will establish ground rules for a safe creative space — and we will enjoy the alchemy of what emerges.