From Walden manuscript

Tapping into the magic of Thoreau’s manuscripts

Digital Thoreau intern Emily Aikens is a sophomore at Yale University majoring in English.

In an age when most literary drafting takes place digitally and can be erased forever with a keystroke, there is something awe-inspiring about engaging with a physical record of an author’s thoughts, mistakes, and ideas. Philip Larkin writes that all literary manuscripts have “a magical value and a meaningful value.” The meaningful value refers to how manuscripts give readers additional context for understanding a printed text. The magical, on the other hand, describes the physical properties of the manuscript: the notes in the margins, the ink stains on the corners, and the kind of paper an author chose to use. Observing these qualities with one’s own eyes is what brings a manuscript to life. It’s what allows one to feel like they are in the room with an author, privy to the secrets that go into constructing a masterwork of literature.

This magical feeling was overwhelming as I paged through high-resolution digital images of Thoreau’s Walden manuscripts. Although I had read the print version of Walden several times, physically seeing Thoreau’s handwriting on the manuscript pages made me feel even more connected to him as a writer. Might a stray mark mean that he was absentmindedly trailing his pencil on the page while deep in thought? Did Thoreau’s change from tight to loose cursive a few pages into “Economy” in Draft A indicate that his hand got tired after pages of furious writing? Even more fascinating, however, were the differences between various drafts of his text. Each of the seven drafts provides its own clues to what Thoreau was thinking. More than that, however, Thoreau’s extensive revisions remind writers today what level of care goes into creating a masterpiece. Although Thoreau scholar James Lyndon Shanley, who worked closely with the manuscript in the 1950s and first arranged the leaves in seven drafts, writes in The Making of Walden with the Text of the First Edition (1957) that the “essential nature of Walden did not change from first to last” (p. 6), careful examination of Thoreau’s manuscripts demonstrates that he evolved as a writer and a thinker from 1845 to 1854. We can learn about Thoreau’s evolution from what he chose to exclude from the published version of Walden as well as from the words that he actually submitted to the printer. Although one can walk into any Barnes and Noble and pick up a modern edition of Thoreau’s published book, that book would not give a full picture of Thoreau’s two-year experiment at the pond.

The history of the Walden manuscripts is complicated. Although Shanley identified seven different drafts of Walden, which the Huntington Library has labeled A-G, none of these drafts is complete. Draft A is the most complete, and subsequent drafts build on what Thoreau wrote in this first draft. While the manuscripts sat largely untouched for many years, Ronald E. Clapper’s 1967 UCLA dissertation “The Development of Walden: A Genetic Text” traces the various drafts of Walden and analyzes the implications that these drafts have for the modern analysis of the text. Clapper is one of several scholars who have attempted to understand what Thoreau’s revisions can tell us about the meaning of Walden. The most recent of these is William Rossi, who uses the manuscripts to illuminate one of Walden‘s most important passages, the description of the thawing sandbank in the chapter “Spring.”

While Clapper’s work throws light on questions concerning the content of Walden, there are still magical qualities of the manuscripts to uncover, namely how the physical properties of the manuscripts affect our understanding of the way Thoreau worked. One of these mysteries is the parallel lines that appear in pencil throughout the seven drafts. Although there are countless markings throughout the manuscript pages of Walden, most of which come from Thoreau himself, many are from when Franklin Benjamin Sanborn, who produced editions of Thoreau’s letters and other works, marked the manuscripts with ink, graphite, and colored pencil while analyzing them for William Bixby, who owned the manuscripts in the early twentieth century. The frequency and range of these markings complicate any attempts to discern their meaning. However, the consistency of the parallel lines conveys that these lines had some significance for the publication process of Walden. Usually coming at the end of a sentence or at the break between two words, these parallel lines do not have any obvious purpose like some of Thoreau’s other pencil markings, such as his notations to reposition passages and revise wording (see example below). Still, given that other pencil markings on the manuscripts are clearly made in Thoreau’s handwriting, without further investigation it is impossible to rule out that Thoreau made these markings.

Bottom portion of manuscript image 357 showing a variety of Thoreau's markings

Bottom portion of image 357, from Draft B of the Walden MS, showing a variety of Thoreau’s markings. The passage is from paragraph segment 20b of the chapter “Where I Lived, and What I Lived For.”

Portion of manuscript image 290 showing parallel lines in pencil dividing a line of Thoreau's writing

Portion of manuscript image 290 showing penciled parallel lines dividing “rubbers over calf-skin.” from “you are no doubt”.

With this question in mind, I embarked on a project to examine instances of parallel lines like those shown above across the seven drafts in HM 924. While I had access to physical copies of some of the manuscript leaves at Yale University’s Beinecke Library, the Huntington Library’s digital collection allowed me to examine each manuscript page in high resolution. I began with a spreadsheet prepared by Digital Thoreau editor and Thoreau Edition Editor-in-Chief Beth Witherell containing instances of parallel lines she had found so far. Beth’s partially tested hypothesis was that these markings correspond to turnovers in the proof sheets that the publisher of Walden, Ticknor and Fields, sent to Thoreau as the manuscript was being prepared for publication. Starting with Draft B, which appeared to have the most instances of parallel lines from Beth’s initial survey, I recorded in a spreadsheet the manuscript image number of each page that showed parallel lines. I then noted the location of parallel lines on the page and transcribed the words on either side of the markings. After going through this process for each of the drafts, I matched the text surrounding the parallel lines to a location in Walden: A Fluid-Text Edition, which presents the information in Clapper’s dissertation in an accessible digital format. This information allowed me to compare the text on each manuscript page to Walden as published. Once I completed this process for each of the instances of parallel lines, Beth checked my work and marked which instances she thought might correspond to proof sheet turnovers. Paying special attention to these likely cases, I went back through all my recorded instances of parallel lines and checked them against the proof sheets. If the text from a specific manuscript page made it into the published version of Walden, I found the location of that text in the final proof sheets that Thoreau sent to his publisher. For the text that did not make it into the published version, those words did not appear on the final proof sheets.

The information I assembled in this way would appear to confirm Beth’s hypothesis. For example, on manuscript image 1254 we read, “Moreover, if you are restricted in your range by poverty, if you cannot buy ∥ books & newspapers.” This section of Walden corresponds to the proof sheet numbers 117 and 118. When one looks at the final line of text on the recto of proof sheet number 117, the last word before the turn of the page is “buy.” The text then continues on the recto of proof sheet number 118 with the word “books” and goes on to finish the sentence. While this instance occurred in Draft F, a similar phenomenon occurs in Draft E. For example, on manuscript image 757 we find,

There is a period in the history of the individual as of ∥ the race when the hunters are the “best men,” as the Algonquins called them.

This line of text comes from “Higher Laws” and corresponds to proof sheet numbers 76 and 77. As was the case with Draft F, the turn of the proof sheet happens between the words “as of” and “the race”—a location that matches exactly with the occurrence of the penciled parallel lines. Ultimately, I discovered that each of the seven drafts (A-G) has at least one instance of these parallel lines corresponding to a proof sheet turnover.

Can we be sure that Thoreau, not Sanborn, made these parallel lines? I brought this question to one of my meetings with Beth and Digital Thoreau director Paul Schacht. One piece of evidence that Thoreau is responsible for them is that there is a marking on manuscript image 905 that says “end of 2nd proof” to the right of the parallel lines and above the word “now.” The text on this image corresponds to the proof sheet turnover from page 28 to page 29 (found on proof sheet numbers 9 and 10, respectively), and the top of the proof sheet containing page 29 has the marking “2d.”, which likely refers to the second version of the proof. Beth noted that the words “end of 2nd proof” appear to be in Thoreau’s handwriting.

Portion of manuscript image 905 showing 'end of 2nd proof'

Portion of manuscript image 905. Thoreau has penciled in “end of 2nd proof” above and to the right of parallel lines.

Top portion of proof sheet showing

Portion of an image from the proof sheets of Walden, HM 925 in the Huntington’s collection, showing “2d.” in the top margin in ink, as well as Thoreau’s penciled instruction to the printer, “Let Shelter be the running title soon.”

In addition to analyzing Thoreau’s handwriting to tell whether he made the markings, we also considered the history of the Walden manuscripts and asked who else could have made the markings. There is no record of Sanborn having access to the proof sheets for Walden, so it seems impossible that he would know which sections of text corresponded to the proof sheet turnovers. In any event, Sanborn was mainly focused on identifying passages that do not appear in Walden as published so that he could incorporate them into his own version of the work, the Bibliophile Society’s 1909 edition. Throughout the manuscript leaves he added notes about material he wanted to copy for or omit from his version. Correlating text in the manuscript with proof sheet turnovers would not have been helpful to his research.

Top portion of manuscript image 290 showing notation in left margin in Sanborn's hand

Sanborn’s penciled notation in the left margin of manuscript image 290, from Draft B, reads “Copy & insert — page 28.” Sanborn’s notation in the top margin reads “omit”.

Given this evidence, it is reasonable to assume that Thoreau made parallel line markings throughout Walden to correspond with proof sheet turnovers. Yet a question remains: Thoreau didn’t mark every correspondence between the manuscript and the proof sheet turnovers, only a selection of them. Why? While sometimes there are over fifty pages between two instances of parallel lines, there are also cases in which parallel lines appear on two consecutive pages. Given this lack of consistency, I began another examination of each instance of parallel lines to discern what their significance might be.

My own hypothesis is that Thoreau used them to mark areas of the text that he had heavily edited and wanted to pay special attention to when reviewing the proof. This pattern of heavily edited paragraphs corresponds with the large majority of the parallel line markings that I was able to identify. For example, paragraphs 7 and 8 of “Where I Lived, and What I Lived For” underwent extensive edits from Draft F to the published version. Thus, Thoreau might have included parallel lines in the middle of paragraph 7—which corresponds to the proof sheet turnover from page 91 to page 92 (found on proof sheet numbers 30 and 31, respectively)—to flag this area of text as one that requires special review. A similar instance occurs in Draft E: Thoreau made parallel lines in the middle of paragraph 15 of “Spring,” and these lines correspond to the proof sheet turnover from page 333 to page 334 (found on proof sheet numbers 111 and 112, respectively). Thoreau also made significant edits to this paragraph between Draft E and the published version.

This theory also accounts for why some parallel lines do not correspond to proof sheet turnovers. While I found 49 instances of these parallel lines throughout Walden, only 25 of them correspond to proof sheet turnovers, which leaves the question: Why are these parallel lines present in the manuscript leaves if not to indicate proof sheet turnovers? One explanation is that Thoreau’s marks intended to identify content for deletion sometimes resemble parallel lines. An example, on manuscript image 383, is the two roughly parallel lines following the word “owl”—one through a comma and the other through a dash. (Thoreau also adds a period after “owl” and revises “and” to “And”.) On a larger scale, some of these areas of text were so heavily edited that they do not appear in the published Walden at all, which is why there is no proof sheet that corresponds to the text surrounding the parallel lines.

Portion of manuscript image 383 showing two lines used by Thoreau to mark content for deletion

Portion of manuscript image 383 showing two lines used by Thoreau to mark content for deletion.

Of course, there is no way to be sure whether Thoreau actually made these lines, nor can we definitively know their function. While some may look at this limitation as a disappointment, I found this lack of certainty exciting. By not having all the answers about the production of his text, we are left to continue wondering about Thoreau and his creative process. Treating the manuscripts as a mine from which there was more knowledge to uncover made my analysis of the text feel relevant and meaningful; it gave me a sense of active engagement with the text that I had not experienced as deeply when I simply read the published version of the book. In other words, this project helped me to tap into the “magical” value of Thoreau’s work.

Not only that, but the unanswered questions surrounding these parallel lines are a reminder that Walden remains a living, breathing text that scholars still do not fully understand. For me, working with Paul and Beth in a collaborative effort to determine the significance of the lines brought Walden to life in a new way and opened my eyes to the scholarly work that one can do in literature other than analyzing the content of a text. As Thoreau writes in “Reading,” “[m]ost men have learned to read to serve a paltry convenience . . . but of reading as a noble intellectual exercise they know little or nothing; yet this only is reading, in a high sense, not that which lulls us as a luxury and suffers the nobler faculties to sleep the while, but what we have to stand on tiptoe to read and devote our most alert and wakeful hours to.” Although the process is tedious, if one makes the effort to “stand on tiptoe” while reading Thoreau’s drafts of Walden, they will undoubtedly uncover what Phillip Larkin identifies as the “magical” element of examining manuscripts.

A new take on Thoreau’s deep cut

I’m excited to announce the latest addition to Digital Thoreau: an electronic version of William Rossi’s essay “Making Walden and Its Sandbank,” which first appeared in print in the Concord Saunterer Vol. 30 n.s. (2022), 10–58. Rossi is co-editor of two volumes of Thoreau’s Journal in The Writings of Henry D. Thoreau, published by Princeton University Press, and (with John Lysaker) of Emerson and Thoreau: Figures of Friendship (Indiana UP, 2009). He also edited the Norton Critical Edition of Walden, Civil Disobedience and Other Writings (3rd edition, 2008).

Making Walden and Its Sandbank revisits one of the most famous passages of Walden—Thoreau’s description, beginning in paragraph 6 of “Spring,” of the “forms which thawing sand and clay assume in flowing down the sides of a deep cut on the railroad through which I passed on my way to the village”—in light of what the Walden manuscript reveals about just when and how Thoreau first drafted, and subsequently revised, the passage. The essay represents an important new contribution to our understanding of both Walden’s and Thoreau’s development. As explained in the abstract to the Saunterer version of the essay:

Close and contextual analysis of Thoreau’s early expansion of Walden’s famous sandbank passage provides considerable insight into his shift toward greater literary and scientific naturalism. Together with its subsequent revisions, this passage, drafted in the spring 1848 Journal six or seven months after he left the Pond, already comprises the core of the narrative’s universally celebrated climactic epiphany. Although scholars have long known of its existence, when its revisions are dated and examined in the contexts of the Walden manuscript’s development, Thoreau’s practice of using the Journal to compose Walden material, the contemporary evolution debate, and his engagement in what historians call “public science,” this remarkable event can be seen both to illuminate and to complicate Thoreau’s career-defining “turn to science” in the early 1850s.

After Rossi generously agreed to republish his essay here, Digital Thoreau editor (and Writings Editor-In-Chief) Elizabeth Witherell and I worked with him to develop a web presentation of his content designed to improve the readability of his manuscript transcriptions and allow readers to compare those transcriptions directly to the corresponding manuscript images in the Huntington Library’s digital collection. To embed the Huntington’s IIIF-compliant images in the essay, we used the community-developed, open source Universal Viewer.

Our collaboration with Rossi led to additional refinements of the essay, as the three of us took a deep dive into the manuscript’s interlineations, cancellations, marginal notations, and “overwritten” stretches (where Thoreau replaced words or letters by writing new ones directly on top of them) and gained additional insight into some of Thoreau’s changes. Preparing the essay for web publication also provided an opportunity to tighten the terminology used to describe the various stages of composition represented on the manuscript.

It’s worth comparing Rossi’s detailed analysis of the sandbank passage’s evolution with the revision model presented in Digital Thoreau’s fluid-text edition as a measure of how much there is still to learn about the timing and content of Thoreau’s revisions, and about the relationship between Walden and Thoreau’s other writings, especially the Journal. “Making Walden and Its Sandbank” will undoubtedly provoke new conversations about Walden’s development, Thoreau’s revision process, and the meaning of a passage central to any understanding of both the book and Thoreau’s thought. It should also bring renewed appreciation for the Huntington’s decision to digitize and freely share one of its most precious holdings and an invaluable resource for both scholars and the general reading public.