Manuscripts Etc.


Ms Homepage Header

The following items are drawn from the William Faulkner Foundation Collection and the Albert Erskine Papers at the University of Virginia's Albert and Shirley Small Special Collections Library (
Click on any image to see an enlargement.

The Mansion

Faulkner began "Father Abraham," as he originally called his Snopes novel, in the late 1920s. By the late 1930s he had decided to turn the novel into a trilogy, and in 1940 published The Hamlet as its first volume. Another fifteen years went by before he began writing the final two volumes, publishing The Town in 1957 and The Mansion in 1959 - more than three decades after writing the "Father Abraham" manuscript. Over the course of this long gestation the basic arc of Flem Snopes' story remained the same, but many narrative details did not. Saxe Commins, the editor with whom he worked on both The Hamlet and The Town, was troubled by the discrepancies he found between these two volumes, and recruited James B. Meriwether, a professor at the University of Texas and a Faulkner aficionado, to help him identify the inconsistencies and reconcile at least some of them. Commins died in 1958, before Faulkner began sending the typescript of the trilogy's third volume to Random House.  His new editor there was a man named Albert Erskine.  As soon as Erskine read the novel's first two chapters in late January or early February, 1959, he began a long campaign to get Faulkner to make the trilogy more consistent. Here you can see some of the documents from that campaign.

Erskine was even more troubled than Commins about the discrepancies, at one point writing Faulkner that he wasn't sure they could advertise the novels as a trilogy when, to cite one example he called to the author's attention, Houston's first name is Zack in The Hamlet, Jack in The Town, and Zack again in The Mansion. Below left: Erskine's first list of discrepancies, sent to Faulkner in late January or early February, 1959. Next to that, Faulkner's reply, in which he expressed a tepid willingness to address these kinds of concerns, but takes his stand on the higher ground of "truth" rather than "fact." Below, center right: Faulkner's next letter, in which he offers a more considered reply to Erskine, who has apparently sent him an expanded list of discrepancies.  Although Faulkner acknowledges two of the inconsistencies that troubled Erskine the most - that in The Mansion Houston is killed riding a stallion he himself had killed in The Hamlet, and that Clarence Snopes' ages in the various books are at odds with each other - he suggests a 'solution' that would allow him to defer revising the texts indefinitely: 'fixing' the novels that had already been published when they are reprinted, which (he adds, perhaps slyly) may also pay off financially for Random House. Erskine, however, persisted. Three months later, after adding new items to the list in discrepancies, he was still trying to get Faulkner to make changes in The Mansion before its publication. Faulkner's reply below right, which Joseph Blotner rightly calls "irascible," takes a stand on even higher ground: that the inconsistencies in the three parts of the trilogy are signs of creative growth, that "alteration" equals "evolution." In the letter Faulkner suggests publishing a version of this credo as a kind of prologue to The Mansion, which Random House did, but that arrangement didn't stop Erskine from continuing to prod Faulkner to make revisions. Like Commins, he engaged Meriwether to help identify the problematic details. A riding accident forced Faulkner to miss a meeting with these two men in the spring, but as the last paragraph of this letter indicates, he did arrange in May to go to New York to work with Erskine in person while the book moved toward publication.

Hamlet Town Mansion Discrepancy List       Letter, Faulkner to Erskine, c9 February 1959       Letter, Faulkner to Erskine, c10 February 1959       Letter, Faulkner to Erskine, c7 May 1959

Erskine was able to persuade him to make at least some changes at this face to face meeting, as is made clear by the note he wrote on one of the galley proofs for the novel (below left). On that galley Erskine's comments are the ones written in purple ink in the margins and typed at the bottom of the sheet; Faulkner's replies and revisions are in blue ink. In the typed comment Erskine refers to "a new scene" Faulkner wrote while he was "here," i.e. in New York in May. Reading the galleys, Erskine realized that that new scene created a new inconsistency, which Faulkner tries to fix at this point with the simplest possible change: changing two "would"s to two "should"s. The other galleys show how Erskine continued right up to this last possible moment to address remaining discrepancies both within the novel's narrative and between Faulkner's narrative and what Erskine refers to on the back of galley 82 as "active historical events" - in this case, the dates of what Erskine calls "the Spanish war," better known as the Spanish Civil War. Faulkner's responses (in blue ink and pencil) show him accepting all Erskine's suggestions, though in one case even after Faulkner had changed his original "eight or ten years" to "four or five" on Erskine's advice, Erksine felt the need to change that again, to "three or four" (galley 81). These are only a few of the adjustments Erskine got Faulkner to make at this stage; some of the others are referred to in the editor's typed comments on the back of galley 82.  The six changes you see Faulkner making here had to occur in July 1959, and before the 29th, which is when he returned the revised proofs to Random House. 

Galley proof 9, The Mansion       Galley proof 81, The Mansion       Galley proof 82, The Mansion       Verso, galley proof 82, The Mansion

Also included in Erskine's papers is the 3-page document below, a list of 134 named characters in The Mansion. This is very likely the earliest list of 'Faulkner's Characters' ever made. Once Faulkner became a staple of the college classroom and the PhD dissertation, a handful of books were published to provide students and scholars with indexes of his characters. The unconventional page numbers attached to certain characters (268A, 419-A, etc.) suggest the list was made from the typescript Faulkner sent Random House in early 1959. But who made the list - whether Erskine or someone else at Random House - and why, whether it was part of Erskine's attempt to identify inconsistencies - are questions that remain unanswered. What is clear is how much work went into making the list.

Page 1, Mansion Character List       Page 2, Mansion Character List       Page 3, Mansion Character List

The materials below date from the end of 1960, when Erskine made a final attempt to eliminate inconsistencies from the story of the Snopeses. Earlier that year Erskine tried to get his author to make revisions in The Town as Random House prepared a new, paperback edition of the novel; despite what Faulkner had said in 1959 they should do for the "next printing of the TOWN," however, there's no record of him doing anything on this occasion.  By the end of 1960 Erskine was editing a one-volume edition of the trilogy. As his 22 December 1960 letter (below) makes clear, this time he prepared very carefully for the task of getting Faulkner seriously to address the discrepancies.

Letter, Erskine to Faulkner, 22 December 1960

Below are the two lists Erskine mentions in his cover letter. They refer to several different kinds of inconsistencies among the first and last volumes of the trilogy, and within each of those individual texts: in dates of events, in characters' ages, in plot details. Erskine was obviously a meticulous editor. But my favorite moment among the details is at the top of page 2 of List I, where his reference to the young protagonist of "Barn Burning" reveals just how sensitive a reader he was as well.

Page 1, Discrepancy List I       Page 2, Discrepancy List I       Page 3, Discrepancy List I       Page 4, Discrepancy List I

      Page 5, Discrepancy List I       Page 1, Discrepancy List II       Page 2, Discrepancy List II       Page 3, Discrepancy List II

The handwritten notes on pages 1, 1-A, 2-A and 3-A of Erskine's lists were made by Faulkner, and they demonstrate his continuing willingness to resolve the trilogy's inconsistencies, at least the ones he couldn't rationalize to his own satisfaction. Yet other evidence suggests that he ignored Erskine's lists for ten months, until just before going to New York in November, 1961, for another meeting with his editor. According to Blotner, at that meeting "he and Albert began working" on the task of aligning all the details in the trilogy, but it's not clear how much progress they made, and the task remained unfinished when Faulkner died in July, 1962.  As the editors of the Snopes volumes in Digital Yoknapatawpha who have made our own lists of discrepancies and characters and so on, we can empathize with how Erskine must have felt.

SOURCES: Joseph Blotner, Faulkner: A Biography; Blotner, ed., Selected Letters of William Faulkner.

    Citing this source:
Stephen Railton, "Manuscripts &c: The Mansion." Digital Yoknapatawpha, University of Virginia, (Date added to project: 2019)