This text is part of the "Bristol Introductions" series which aims to present perspectives on philosophical themes, using non-technical language, for both the new and the advanced scholar. This work argues that over the centuries both philosophers and linguists have underestimated the extent to which their linguistic discussions have been locked into a metalanguage which they have vested interest in protecting and perpetuating. This metalanguage (which includes such terms as "language", "sentence", "word", "grammar", "meaning", "true" and "false") generates some linguistic problems, while inhibiting the recognition of others. It also tacitly marks the boundaries within which solutions can be proposed and defended. In short, this metalanguage supplies the Western code of conduct by which linguistic debate can be engaged. Thus, the book raises the question of whether metalanguage itself, as a historical product, is not the source of serious limitations in both philosophy and linguistics. This question has its own problems; for to reject the metalanguage entirely is to undermine the disciplinary foundations of both subjects - as well as leaving unclear on what basis linguistic inquiry should henceforth proceed.