This kind of analysis may serve, for instance, to test the authenticity or authorship of a text. In much biblical or Shakespearean scholarship, the received text is compared with possible sources, or competing versions are compared with each other, in order to establish a canonical text. Editorial overlays are peeled away to reveal the presumed original text. If authorship is in doubt, stylistic comparison with other texts is used to support arguments for or against various contenders. Where the purpose for which a text was edited is known, analysis of input and output can establish whether that purpose has been achieved. If a text has been edited for a particular reader- ship, comparison with the original can show whether the edited version is in fact more suitable for the readership than the original. Texts are often re-written for learners, whether foreigners or children, according to read- ability formulas. The study by Davison et al. (1980) indicates how writing to such formulas does not necessarily produce a text suitable for the target readership.