Editorial Policies
Written by Steven Amarnick (last edited 12-29-20)
I use the collective “we” to reflect the countless hours of discussion (most of them with Robert F. Wiseman in all phases of the project; many with Susan Lowell Humphreys in earlier phases and Michael Williamson in later phases) that led to the policies. All the final policies, however, were mine. For more about the genesis of the project and the various phases, see my Introduction to the Folio Society edition, elsewhere on this website.
Our purpose was to produce a version of The Duke’s Children as close as possible to the first edition (FE) of the book that Trollope would have expected to publish under ordinary circumstances—that is, when he was not forced to make substantial cuts to save space. One important exception, however, is that we deferred to the manuscript (MS), rather than the Chapman & Hall first book edition (FE), in punctuating dialogue and letters. We were also willing to correct sloppy errors in certain circumstances that made it into the FE and have been published until now, and we have made spellings and capitalizations consistent in a way that neither Trollope nor his publishers ever did. There are, of course, complications. For instance, when Trollope in 1878 looked at the text for the first time in two years, with the aim of shortening it for the All the Year Round (ATYR) serial publication, he was mostly focused on cutting the text yet still making sure that it flowed smoothly—something he achieved in remarkable fashion. Nonetheless, it seems clear that at times, as he took a fresh look at a passage, he made editorial changes that had nothing to do with saving space. Should we respect his wishes and print those changes, even though he wouldn’t have made them if he had published the book in his more customary manner? The answer almost always is yes, but we also recognize in some cases that his distance from the text might have made him understand it differently than he would have in 1876, sometimes to the detriment of the novel.
In comparing the MS to the published novel as we determined what to print, we used the Chapman & Hall FE rather than the ATYR first periodical publication because the former’s punctuation is much closer to Trollope’s. In addition, his crucial 1878 letter, written just months after he cut The Duke’s Children, makes it clear that he considered the book to be more authoritative than the periodical. In that letter (which I discuss in my Folio Society Introduction), he expresses a willingness to cut his novel John Caldigate for the periodical, but only if a clean copy of the full text could be preserved and then printed in book form. No doubt he had learned his lesson from The Duke’s Children, where his having marked the cuts directly on the MS made it too difficult to reconstruct the full text for Chapman and Hall.
In the document below, we explain our guiding principles. For explanations of many specific applications in each chapter, see Notes on Editorial Decisions elsewhere on the website.
Straight-Line Versus Wavy-Line Cuts
Trollope wrote The Duke’s Children in 1876 and shortened it in 1878. We believe that when he used a wavy line to cut a word or several words, he did so in 1876 as he was writing—immediately crossing something out and replacing it with something else. It is possible, too, that he made some of those wavy-line changes the day after he first wrote the words. (Trollope talks in his autobiography of spending the first half-hour of every three-hour morning session re-reading the previous day’s work.) What this means is that, with just a few exceptions, we did not restore anything that was crossed out with a wavy line.
This practice on Trollope’s part applies only to small cuts. For, say, entire paragraphs, Trollope did use a loose wavy line—or, often, a large X.
In a few cases, it is difficult to say with certainty whether a cross-out is straight or wavy. If the word underneath was decipherable, we sometimes restored it and sometimes did not—depending on the context.
Occasionally, Trollope made some regular (as opposed to space-saving) changes with a straight line—either in 1876 while reading the previous day’s work, or in 1878 while cutting primarily for space. In most of these cases, words cut with a straight line were replaced with longer words—so that, obviously, the change did not shorten the text. There are some instances, however, where we believe that he was not aiming to save space, even when he used a straight line, and even when the replacement was shorter. In those instances, we accepted the cut. To reject a straight-line cut and call it non-space-saving, we needed to be reasonably certain.
We believe that Trollope was willing to cut a word and replace it with another to delete even one character—as long as he found the change easy to make. What we do not believe plausible: Trollope replacing a group of words with an entirely different group of words, only to eliminate a character or two. It is unlikely Trollope would have been willing to take the time. In some instances where such a change did occur, it is possible to see why Trollope preferred the new passage. Thus, such changes should be seen as regular editorial revisions rather than space-saving cuts.
Cutting Adjustments or Regular Revisions?
Usually it is clear enough when Trollope made a cutting adjustment, as he inserted new words to replace a space-saving change. In such cases, we left out those insertions when we restored the passage that was cut. Sometimes, though, we kept the insertions at the same time as we restored the deleted passage. We did this when there was a strong case to be made that the new words were put in as regular revisions and not as cutting adjustments.
Errors—both the compositor’s and Trollope’s
We corrected obvious compositor errors, and any egregious grammar errors, along with other errors that Trollope made, as long as those corrections could be done simply—without having to remove or rewrite whole phrases or sentences. Certain other errors that occur in dialogue were also allowed to remain if there is any chance that the character could be construed as making the error.
In the few cases where we believe we were justified in tinkering with the text, we made the least intrusive moves possible.
Though we restored nearly everything that Trollope, for space reasons, cut with a straight line, we recognized some situations where he almost surely would have made a slight change if he had seen the original sentence in proofs. We always looked for plausible reasons to restore a word, but if it seemed too obvious that Trollope erred and would likely have made the correction himself (or would have been glad to make the correction if someone had alerted him to the need for one), we made the change.
It is evident that Trollope did not read proofs of The Duke’s Children against his manuscript, and that he could be a hasty proofreader. Thus, if there is good reason to believe that the compositor made an error, one that Trollope did not catch, we reverted to the MS. There is certainly the chance, in some of these cases, that Trollope made the change in proofs, but we think it unlikely.
If the FE prints a word that looks very similar to the word in the MS, our first assumption was that the compositor made an error—though we recognized other possibilities as well.
Changes Made in Proofs
Plenty of tinkering was done on the text beyond the MS. It is hard to know when these changes were made by Trollope himself, or, say, the printer’s reader—though even in the latter case, one could claim that Trollope gave at least tacit approval if he allowed the change to stand. We printed most of these changes—unless there was some compelling reason to reject them. There is no convincing evidence that, in at least a huge majority of these situations, he made the changes to save space. And even if, occasionally, it is possible that Trollope was still looking to cut, it would be difficult to distinguish between space-saving cuts and those made as part of ordinary polishing.
We rejected the FE text if the changes were likely caused by the particular circumstances of the cut passages. If we believed that the change was the result of a space-saving cut nearby—or even, in some cases, a cut made in another part of the book.
We also rejected the FE text if we believed the change was a result of compositor error, or, occasionally, Trollope in 1878 misunderstanding what he had done in 1876. We were cautious about resorting to either explanation, but if a change weakened the text, we sometimes disregarded it, depending on the degree of harm.
Close Calls
Trollope’s handwriting is difficult—indeed, this is the last manuscript that he wrote entirely in his own hand. In some cases, we were faced with two plausible choices—and occasionally chose a word that seemed slightly less likely, based on the strokes involved, if we felt that the word fit the context better.
We flag (Notes on Editorial Decisions) all word readings that we had a high degree of uncertainty about. We also note instances where we could not decipher a word at all. In such cases, we put a dummy word in its place if absolutely necessary. If the sentence reads fine without the word, however, we left it out.
Punctuation—General Principles
We maintained Trollope’s MS punctuation in letters and dialogue, and we used FE punctuation in narration. There are too many awkward aspects of the MS narration punctuation to use it everywhere; we can indeed see how the FE punctuation sensitively (without too many changes) cleans it up—and why Trollope would be appreciative. (The changes in ATYR are more radical, for instance removing all his comma-dashes and semicolon-dashes.) With dialogue punctuation, however, it is remarkable how few errors Trollope made, other than leaving out a number of necessary question marks (or occasionally putting them where they don’t belong). Trollope was especially proud of, and attentive to, his dialogue, as he makes clear in several pages of his autobiography. We believe that Trollope’s dialogue punctuation represents the fine-tuned way he heard the words being spoken (with sometimes ever so slight differences between commas, comma-dashes, semicolons, semicolon-dashes) and thus should be preserved whenever possible.
In following MS punctuation for dialogue, we still deferred to the FE when there was any ambiguity (if, say, it was hard to tell if Trollope had written a comma or a semicolon), or when the change was likely made in the FE to prevent a reader stumble. Indeed, in cases involving a potential reader stumble, we made minor punctuation changes even if both the MS and FE had the same mark. We also ignored a punctuation mark, or put one in, if it was very clear that the MS or FE punctuation mark (or lack of a punctuation mark) changed the meaning of a sentence in a way that Trollope could not have intended—in other words, if the punctuation was clearly problematic.
If in dialogue a punctuation mark is somewhat awkward, and there is even a slight chance that it could be a blot or stray mark, we were willing to leave it out or make another change.
Though we followed Trollope’s MS punctuation for dialogue, we recognize that there are surely cases in proofs where he would have changed the punctuation himself. It is impossible to identify every such case, but in certain egregious or obvious instances we followed the FE punctuation—with the strong sense that Trollope indeed did make the change.
We accepted idiosyncratic punctuation as long as it was not excessively distracting (acknowledging that different people have different concepts of what is distracting).
We recognize that, while Trollope did allow Chapman & Hall to alter his punctuation, there are occasions where the publisher erred—and that Trollope did not catch those mistakes in proofreading. There are also times that Trollope erred and Chapman & Hall merely followed Trollope’s punctuation. We corrected the most glaring of such errors.
If the FE changed more typical punctuation to something highly unorthodox, we deferred to the MS.
In cases where the MS punctuation is clearly superior, we looked for reasons to restore it: perhaps something about the handwriting that might have made the compositor err, or something about the surrounding cuts that changed the circumstances of the punctuation.
Though we followed Chapman and Hall’s FE punctuation for narration, we recognize that the different sound of altered sentences may have influenced their decisions. Thus, we reverted to the MS punctuation for restored sentences in cases where it seemed especially warranted.
Punctuation—Specific Usages
Given how often Trollope forgets to put them in, we allowed ourselves the freedom to add question marks even in cases where it is possible (but not likely) that the sentence could be read as a statement rather than a question. The opposite is true too; some question marks should be changed to periods. However, we only made such changes in cases where readers might have stumbled.
Because it is too distracting otherwise, and because the FE almost always does include commas in these situations, we added commas before –ing verbs when more than just one or two words follow.
In dialogue, we added commas around “you know” even when not in the MS.
We used the serial comma for narration, even if occasionally it is missing in the FE, but we accepted its inconsistent use in dialogue.
In following MS punctuation for dialogue, we recognize that dashes could be hidden under straight-line cross-outs. Thus, if the dash fits, we assumed it is there in relevant situations.
Occasionally, in broken-off speech, Trollope also includes a comma before the double dash. We kept that comma, as it likely represents Trollope’s equivalent of ellipses—a pause before the broken-off speech.
We used double dashes and no period when there is broken-off speech at the end of a sentence.
When there is a dash at the beginning of resumed dialogue, we put the dash outside the quotation.
We added commas when necessary in non-restrictive clauses, even when they are not in the MS dialogue or the FE narration.
We added a comma before and after a person’s name during direct address.
In cases where a balancing dash is clearly appropriate, we added one. However, we left out the second dash if there is a lengthy distance from the first dash and no likelihood of a reader stumble.
Sometimes Trollope—or more often Chapman & Hall, ignoring what Trollope wrote—used plain dashes rather than comma-dashes. We used comma-dashes when they seemed more appropriate.
Spelling, Capitalization, and Italics
We used the spelling and capitalization of the FE. However, the FE–and subsequent OUP editions of the novel, which follow the FE closely—are often inconsistent, in which case we had some flexibility about which spelling or capitalization to choose.
We spelled out numbers, unless the number is part of a letter that a character is writing, or unless the number is over a hundred and contains more than two digits (“eighty-eight,” but 188; five thousand, but 5001).
We italicized foreign words.
Paragraph Breaks
We deferred in most cases to the MS when a paragraph break appears in the FE but is not indicated in the MS. Many of these breaks were instigated by the cuts, and many of them were made in error, after Trollope crossed out a number of lines within the middle of a paragraph.
A Brief Note on Editorial Pragmatism
Rules and guidelines are important, and we followed them—sometimes, as with punctuation, obsessively. Yet we also left ourselves some wiggle room. There are many places where we followed rules even if they did not lead to our preferred result. And there are also many places where we got to the result we wanted—as long as we could come up with a reasonable justification. We explain our judgments in Notes on Editorial Decisions.
