A neurology of belief

The notion that all mental acts, all mental processes and dispositions have specific neural correlates has become much easier to explore in the past 15 years with the development of PET scanning and especially functional MRI. We can now, for example, demonstrate activity in the visual cortex when a subject views a test object, and we can pick up similar activity if we ask the subject to imagine or make a mental picture of what the object looks like. Functional brain imagery has also been used in relation to more complex mental processes, such as those involved in economic decisions. There have, however, been no comparable studies addressed to the neural correlates of belief in general until Harris, Sheth, and Cohen’s pioneering article in the present issue of Annals of Neurology. Harris et al.’s experimental method, both simple and ingenious, was to develop a battery of statements which were presented in written form to subjects while they were in the fMRI scanner. The statements in seven different categories (autobiographical, mathematical, geographical, religious, ethical, semantic, and factual) each were designed, according to the authors, “to be clearly true, false, or undecidable.” In the mathematical category, for example, the statements were: