Which comes first – passion for the scientific question or passion for the organism? For most biologists I think it's the former, but for me it was the latter. I became smitten with honey bees at the age of 18 and have never looked back.

Once immersed in study, the question did come: how can a honey bee, with a brain the size of a grass seed, create a collective organisation in which all tasks are divided efficiently but flexibly among as many as 50 000 individuals? Honey bee division of labour is a spectacular example of social behaviour; trying to understand its mechanisms and evolution has motivated most of the research in my laboratory over the years and also has led periodically to rewarding expeditions into new scientific terrains.

After starting with behavioural and endocrine analyses as a graduate student at Cornell University with Roger Morse, my postdoctoral studies with Robert Page at Ohio State University demonstrated for the first time heritable influences on division of labour (Robinson & Page 1988). Then mechanistic studies in my own lab at the University of Illinois revealed striking differences in brain chemistry and brain structure between bees performing different jobs, raising the possibility that these changes were orchestrated by changes in brain gene expression (Withers et al. 1993). To enhance our ability to discover insights into the mechanisms and evolution of this form of social behaviour, I decided in the mid 1990s to initiate a molecular component to our research programme.