This paper presents a purely pragmatic account of quotation which, it is argued, will be able to accommodate all relevant linguistic phenomena. Given that it is more parsimonious to explain the data by reference to pragmatic principles only than to explain them by reference to both pragmatic and semantic principles, as is common in the literature, it is concluded that the account of quotation presented here is to be preferred to the more standard accounts (including the alternative theories of quotation, discussed here).