My paper 'The Use of Spellings for Compositor Attribution in the First Folio' was published in the March 2016 issue of Papers of the Bibliographical Society of America. You can download it below.

Readers of the published version of the paper who are looking for the introduction to statistical concepts which is promised there may jump straight to that section.

The version of my paper provided below is the one that was peer-reviewed by the reviewer chosen by PBSA, and revised by me to take account of the peer-reviewer's comments, but before it was copy-edited for publication. It is my preferred version of the paper and it differs from the published version in the following ways:

• The section called 'Statistical Analysis' was abridged in copy editing, omitting the introductory passage aimed at beginners which is included here.

• The peer-reviewed version of the paper included a section called 'General Theories of Compositor Analysis' which is missing from the published version. The journal, but not the peer-reviewer, objected to it, regarding it as a serious misrepresentation of one of Hinman's early papers. I was unwilling to accept their suggested changes to it and asked them to omit it. It is included here.

• The 'Conclusion' section was abridged for publication, consistent with the omission of the section on general theories. The original peer-reviewed version is given here.

• Many changes of wording and a few other abridgments were made in copy editing. The version given here contains the original peer-reviewed text. It also places the tables at the points where they are discussed; this was impossible to achieve in the published version owing to typesetting constraints.

After proof-reading the published version I discovered an error in Hinman's Printing and Proof-Reading which I had assimilated into the work that led to the writing of the paper. At II,277 Hinman stated that rr1r is an Ex page and rr6r a By page. That is consistent with the compositor attributions he gave at I,393 and with the ones given later in the Norton facsimile. However, at II,517 Hinman accidentally transposed the data and wrongly gave rr1r as a By page and rr6r as an Ex page. For my work I took Hinman's attributions from II,514-18 and thereby assimilated his error. The error was of very little significance because in almost all of my paper I use the attributions from the Norton facsimile rather than Hinman's attributions. The only parts of my paper that were affected were Table 6, Table 17 and the lists of spelling preference reversals for Hinman's attributions given in the Appendix. For the sake of accuracy I have made the necessary corrections in the version given here but no conclusion drawn in the paper was affected by the error.