The Blake Archive is going to publish the first 40 years of printed back issues of Blake/An Illustrated Quarterly (1968-present, so those 40 years take us through 2008), which is now published online on the Open Journal Systems platform. The issues have already been scanned and OCRd, proofed and corrected against the originals, and encoded using TEI Lite. One of the latest things we’ve been trying to figure out in the Blake Archive’s back rooms is what policies should govern our use of published Blake Archive images in the first installment of back issues, which we hope to publish later this year or early next year.

As usual, an apparently simple editorial question turns out to be a mare’s nest that calls for a policy, or policies, plus unpoliced leeway for cases that the policies can’t quite reach (the leeway is in the territory that I call xediting).

The images in the quarterly, as you might imagine, have always been mostly mediocre black-and-white halftones. The Archive’s images are, by comparison, glamorously better–higher resolution, color (and accurate color at that). So it makes sense in publishing the old issues to substitute the Archive’s better images for the quarterly’s old doggy ones. When we can make a one-to-one switch it’s easy.

But what constitutes, say, an approximate equivalent substitution for images that we have, shall we say, versions of but not one-for-one replacements? And can we draw on images that we have stored on the work-in-progress site that the Archive maintains behind the scenes (where we have a huge bank of unpublished images), or do we have to stick to the images we’ve actually published in the Archive?

And when we make substitutions, should we say so? And if we do, how much should we say? Something simple, like “substitution,” or something more specific to the case at hand? And what about correcting mistakes that we inevitably uncover as we work through these illustrations and captions published years and years ago? Should we make the corrections or leave the original error intact in the republished version? Or leave it intact and note the error? Or . . . ? In other words, how faithful should the newly edited online version be to the original?

The best decision we made back when we were first planning to publish the back issues of the quarterly in the Archive was to preserve a PDF of every back issue and to make those PDFs available to all users of the back issues through obvious links. So the original, in all its no-doubt error-riddled glory, will be there for any researcher to consult at any point. That also frees us up to correct the new-made Archive version whenever we feel it’s useful to do so. (I see this as another xediting moment: the decisions we make about when to correct and when not to correct will inevitably be partial and inconsistent–will depend a lot on attention span, available resources, etc., which always play a major–I’d say fundamental–part in editorial activity.) Enough for now. We’ve got to make those decisions and push those back issues out.