One of the most characteristic features of geolinguistic study is the recourse to maps and diagrams; authors often supplement words with illustrations of patterns and processes stemming from the spoken word. This transposition of word and image is significant, for maps and diagrams are a sign of language. The concerns of the linguistic map maker can be seen to mirror the more general debate about methods in the social sciences, and particularly those in the developing discipline of geolinguistics. It is contended that practioners of geolinguistics have not clearly recognized the role and value of illustration, nor have they been able to exploit it to the full. The same can be said of illustrations as of words: used without a full understanding of their meaning, they can confuse rather than clarify the message. Accumulating evidence suggests that maps can make useful servants in linguistic study, but that they are not without the potential to disappoint and even deceive their users and makers. As with all other means of imparting information, linguistic cartography depends on good quality data. (JL)