In this book, I have largely been re-working old ground, in the sense that I have only dealt with the representation of a very limited vocabulary, namely, that which standard first-order logic aims to represent. The main burden of the book is that the standard representations are unsatisfactory on certain counts and must be modified – especially the ways in which count nouns and proper names are represented. But the purpose of the exercise has been to prepare the ground for representation of a wider vocabulary and especially of expressions which first-, or even second-order logic is effectively unable to handle. I should like to leave the reader with at least a foretaste of these possibilities, and there is one topic which allows me to do so in a brief compass: the representation of adverbs. This will also provide an occasion to introduce a final structural innovation: one which, I believe, will eventually have large consequences.

Adverbs are awkward for Fregean logic because they appear to qualify verbs, and it is clearly out of the question to represent them by means of conjunction in the way that is plausible for some adjectives. Thus it is obvious nonsense to try to analyse ‘Mabel spoke slowly’ as *‘Mabel spoke and Mabel slowly’ in a way that it is not obvious nonsense to try to analyse ‘Diana is a brown cow’ as ‘Diana is a cow and Diana is brown’.