Coupling documented artifacts and methodological reflection, this paper will center around cyphering books (alternative spelling ‘ciphering books’) from archives of eighteenth- and nineteenth-century North American education. It will present excerpts from these hand-written mathematical exercise books and reflect upon the productive paradox that the cultural historian faces when studying them. On the one hand, these books often testify to the idea of officialism and longevity; they may be several hundred pages long and exhibit rather rigid genre elements (verbatim noting what the schoolmasters read out; adhering to calligraphic precision). On the other, cyphering books may also document social history, as when the exercises of simple and advanced arithmetic are enriched with personal marginalia, including doodles, puns, poems, spontaneous penmanship exercises, and invectives against other individuals in the classroom.