Research articles and patents are claim-making texts that work in diff
erent social arenas. They have in common a process of enrolment, by wh
ich diverse interests are incorporated textually to support a claim. B
ur the evaluation of a patent draws on different assumptions about the
nature of a claim, the relation to other texts and the account of fut
ure effects. One way to bring out these differences is to look at the
experience of researchers who are used to writing research articles, a
s they learn to write their first patents. I follow the drafting of pa
tents by a zoologist (applying in the US) and a medical researcher (ap
plying in the UK). They revised their texts in response to the comment
s of their patent agents, the patent examiners, the potential sponsors
, and in response to their own continuing laboratory experiments. In b
oth cases, the crucial areas of debate were the scope of the claims, t
he relation to prior and competing texts and the story that linked dis
covery and a range of possible applications. in each case the research
ers and their agents developed textual devices appropriate to the disc
ourse of patents: nesting claims, constructing an ideal reader, doubli
ng the story. Their strategies for translating interests between their
laboratories, sponsors, the patent agencies, potential users and the
wider public raise the question of how separate arenas of knowledge an
d ownership are defined.