Autonomic neural adjustments to the challenges of real life generally occur not as altered activity in a single outflow but as patterns involving a number of outflows, usually linked with a behavioural response and often an accompanying emotion. The eminent Swiss neuroscientist Walter Hess shared the 1949 Nobel Prize in Physiology or Medicine for demonstrating this principle. He showed that stimulating different sites in the diencephalon of conscious cats could trigger autonomic responses appropriate for (e.g.) defence, hunger, defecation, sleep etc., and these were accompanied by the corresponding behavioural and emotional responses in the animal1 . These combinations were self-evidently biologically meaningful, and gave rise to the view that a limited number of autonomic-emotional-behavioural ‘modules’ were hard-wired into the brain as pre-formed (though modifiable) motor patterns2 . Hess used electrical stimulation for his studies, however, which would have activated passing and collateral fibres (setting up axon reflex responses) as well as cell bodies. For that reason a modern reassessment of his experiments, while not invalidating his essential findings, raises questions about exactly where are the individual neurons (‘command neurons’) whose activity drives more than one outflow. Those questions have been a recurring topic of discussions with Roger Dampney, so it is apposite to raise them here. In this presentation I will review some of the progress on this issue, examining what we know about what is integrated and what is not integrated at different levels in the autonomic neuraxis, and suggest some possible ways forward.