I’ve spent the day making corrections to transcriptions, and geez it’s difficult. Not corrections, exactly, because what I’m doing is attempting to record the characteristic qualities of individual copies, such as type damage.
I’m working with three copies of Uncle Tom’s Cabin, individual copies of the paperbound Million edition. And the difficulty with this edition is that it was reprinted 10s of 1000s of times, and so in addition to being printed on small type, on cheap paper, and in something of a hurry given the number of uncorrected errors, the type deteriorated rapidly. The last, type deterioration, results is alterations that look like errors, but you can’t really know whether an individual example is or isn’t an error until you check another copy.
I have a double-keyed and oral-proofed digital transcription and two sight collation records based on work with Lindstrand comparators. One collation record stretches 8 pages in smaller handwriting, another 18 pages in larger handwriting, and there are about 150 variants that must be examined in minute detail in 4 different copies.
Of course, what, exactly, is an error? Is it an error only if a person did it? Or does type damage become an error if for the reader it seems like one? My codes are “td” for type damage, “err” for errors, and “vp” for variant printing. But sometimes you just wonder how, exactly, can one decide? And once one starts making decision, then each principle just starts heading toward Alice-in-Wonderland rabbit holes. And what if in the morass of copies one still doesn’t know—and can only make a good-faith effort? The answer to last is simple. If you don’t know, record all problem spots and then query librarians to see if an individual copy in better shape than the one’s available to me. Then after you find that imaginary pristine copy, check that one.
With all the wonders of cheap reprints and OCR, I often realize that readers have no clue about the world of difference between cheap OCR and proofreading, serious editorial effort to transcribe accurately, and fanatical scholarly-editor insanity, in which every jot and tittle in multiple copies must be examined minutely simply because it’s the right thing to do.
But today, after 7 full hours of detail work, examining minutely the broken type and paper tears, thin space before comma that might signal semicolon missing dot or shifted comma, on 4 copies of 60 pages of text, I’ve had it. Had it. Had it. Had it. Pulling out my hair and lost my marbles had it.