In his 1935 book The psycho-biology of language, the linguist George Kingsley Zipf introduced the concept of dynamic philology, which he hoped would integrate the formal and quantifiable aspects of the psychological sciences with the philologist's concern with the social and cultural contexts of speakers, writers, and their linguistic productions. Yet Zipf's modern impact has largely been in large-scale statistical analyses of word frequencies in corpus linguistics and psycholinguistics, while many humanists are rightly skeptical of anything calling itself philology that is divorced from social context. The present paper uses material from the study of numeral systems - a core subject of traditional philology - to propose a different configuration of "dynamic philology".