Their book dismissed much of the work in the field as being 'without scientific value', and went on to argue that single-layer networks could not represent a whole class of problems.