When we talk about the brain, we have to choose between one of two models. When we describe or try to understand anything whatsoever, we do it by likening it to something else we think we already understand better. There are only two models available for the brain: the machine and the person. In the end we don’t have a language specifically for hemispheres. We only have the language we developed for people or for machines. Using the machine model is an approximation: so is the person model. A single hemisphere is capable of sustaining human life – and therefore being involved in the processes of ‘wanting’, ‘aiming’, ‘desiring’, ‘liking’, having ‘values’ – this is no more of a distortion than pretending it is simply a machine. When we see two hemispheres in the same person treating things quite differently – clearly valuing and favouring some things more than others – as can be seen in split-brain subjects, for example, it is almost perverse not to allow one to speak of the hemisphere as at least having some of the qualities of the person that relies on it. The fact is that we don’t know what sort of thing the brain is – or even what a single neurone is. The neurone is often modelled as a wire or a chip: however it is a vastly complex self-regulating partially autonomous system, with tens of thousands of channels and ports interacting with one another and the whole context of the body in which it lies, manufacturing, transmitting – it isn’t fully modelled as a wire or chip. Much is left out. Still less is the whole brain fully modelled as a machine.