Communication between researchers schooled in developing complex climate scenarios and geographic information (GIS) on one hand and local people who possess a wealth of knowledge on local conditions on the other pose serious challenges for research into human adaptation to climate change. Often, researchers who have developed sophisticated scenarios of future climatic conditions or geographic datasets find that they face an array of non-scientists who do not grasp either how the scenarios or datasets were developed or misinterpret scenarios as a prediction of future events. Similarly, local people often possess a wealth of knowledge on micro-climate, soils, timing of rainfall, history of cropping, spaces of production and many other valuable datasets that appear incommensurable to scientists who are unable or unwilling to challenge their own epistemologies. Researchers interested in the potential impact of climate scenarios on human population must contend with not only this first translation problem, but must also translate how local people negotiate not only environmental stressors, but also political, social and economic ones as well.

In this paper, we describe our methods of integrating these two seemingly polar opposite epistemologies in a case study set in rural southeast Malawi. We focus on in-field methods of bridging diverse epistemologies by describing our survey techniques, which include structured household quantitative interviews and informal qualitative group discussions. We devote particular attention to issues of interview bias, gender bias, age bias and local political conflicts. We also show how we tried to avoid leading respondents to particular answers.

In addition to epistemological and conceptual translation, we examine the linguistic translation of climate scenarios and geographic information datasets to populations with low literacy rates and where English is not the first language.

Our initial findings indicate that much forethought is needed to the issues of language, time and space in order to craft studies that integrate diverse epistemologies. Language is a critical element of cross-episteme translation. Even had respondents spoken English as their first language, the vocabulary used in climate-GIS studies in often unapproachable. Finding terms that both convey the meaning intended and translate correctly into Chichewa, the local language, was a primary focus. We also found that concepts involving the use of time need careful attention. Straightforwardly asking a respondent about events in 1998, for example, will not yield robust answers. We coupled an event calendar, where we highlighted major events in the nation, to a local environment calendar, where we identified anomalies in precipitation or temperature that stood out for local people. Finally, we found that local understandings of space, particularly land use, were radically different from those engaged by “experts”, however, the two could be made commensurate through the use of transect walks, mental mapping, and geographic positioning systems.

The hybrid epistemology that emerged from our research will strengthen research into coupled human-climate change in three ways. First, such methods will help increase the relevance of climate scenarios to local groups facing climatic variability. Second, local knowledge can be better integrated into the science of climate change once commensurable methods for obtaining such knowledge can be shown to be effective, and finally, such research challenges conventional and scientific knowledge around the concepts of language, time and space that are desperately needed to address the impact of climate change on humans, and in our case, some of the most vulnerable humans on Earth.