The author is a principal of Windfall Projects, and the publisher and editor of ConcordMA.com, which includes this publication.
Introduction: Henry Thoreau actively took up land surveying in the 1840s. When he died in 1862, he left a sizable ( body of his working papers -- field notes and draft surveys -- in the care of his sister Sophia. The 1874/75 report of the Library Committee of the Concord Free Public Library recorded an important deposit: "Miss Sophia E. Thoreau has deposited in the iron safe of the library building the unpublished manuscripts of her brother, Henry D. Thoreau. They fill three trunks or boxes. One contains a complete survey of almost every farm in town ..." The 1876/77 town report referred to the subsequent bequest of the survey trunk to the CFPL.
The library holds approximately two hundred manuscript Thoreau surveys made between 1849 and 1861 of properties in Concord and elsewhere. A number of parcels are represented by multiple draft versions. (The final surveys ended up with the parties who hired and paid Thoreau to do the work.) A small number of surveys in the library's collection are final versions, received from donors other than Sophia Thoreau.
In 1999, to provide full researcher access, the library began to scan and mount the surveys on its web site. However, it soon became apparent that the project was beyond the capabilities of available resources. Consequently, in 2001 the library sought and in 2002 received $17,500 from AT&T to undertake the project described below.
-- Leslie Perrin Wilson
Special Collections, Concord Free Public Library.
"Wow -- they're so wonderfully clear and legible!" said Leslie Perrin Wilson, Curator of Special Collections at the Concord Free Public Library.
Legible? Clear? This is not the usual first reaction to Henry David Thoreau's famously difficult-to-decipher hand. Yet it is the immediate impression upon seeing the Library's 200 Thoreau land and property surveys now available online. In fact, in the very short time they have been in digital form, the surveys' clarity has already allowed some of them to be better understood than before. You can view the Library's entire collection of these surveys here (click on "Thoreau Surveys").
An Unusual Project with Many Challenges
Because we could not find another similar project already online when we began this one in 2001, we had to develop our own approaches to digitizing and presenting these documents. Whatever we did, it had to be economical, quick, accurate, and gentle to the originals. It also had to render highly-readable output. This article will briefly outline some of these challenges and how we met them.
Our first task was to determine for whom were we preparing the surveys and why. Was the audience the general public? Thoreauvians in general? People directly served by the Library? Academics doing research? Were we aiming for a faithful reproduction of the original documents, an emphasis of the artistic merits of the surveys, or maximum legibility? These considerations were important because they determined the shape of everything which followed.
Curator Wilson decided the main goal was to provide maximum legibility for research purposes. The documents are presented with the assumption that academics use some type of high-speed internet access. A speedy download was deemed far less important than were clearly presenting the richest, clearest images. If you connect to the site using a dialup (like most Concord residents do because of a dearth of other options), you will find the enlarged views require patience to download -- the pages with the thumbnails, however, should be fine for most users.
We decided that each survey would be presented on its own page. Typically, a 3" or 5" version of the survey would be shown along with Thoreau's description. This "thumbnail" could be clicked on to show an enlarged view, which would be slightly over 200% of the original. (Users of Internet Explorer 6.0 should be certain to click on "Survey Printing and Viewing Tips" at the bottom of each survey page for information on how to defeat that browser's tendency to reduce large images to fit on the screen.)
Many of the surveys were too large to present as a single enlarged image; these would be offered instead in sections. Clicking on a different part of the thumbnail would result in a different enlarged image. Over time, we were pleased to find that this plan yielded consistently good results and no changes were required; it is this approach you see on the final site. By the end of the project, we would create and manipulate nearly 1700 text and image documents.
Scanning large documents (the largest in this case being the Concord River survey at over 7.5 feet) is typically done with equipment specifically designed to handle large-scale originals. But this type of equipment is quite costly -- on the order of tens-of-thousands of dollars. In our final analysis, we decided that the labor required to use a high-quality, slightly-larger-than-usual flatbed scanner would be far less costly and would yield images of more-than-adequate quality for our purposes.
Inconsistency, Thy Name is Henry
O, you of the many pens and pencils -- some light, some dark...some heavy and others soft as the gentlest rain. Did it give you pleasure to use all of them at the same time on so many individual surveys?
Did it bring you joy to make your page corners far from right-angled? Did you laugh as you scribbled right up to one edge of a page while keeping acres of untouched snowy whiteness along the other?
Was it a lesson in the Infinite to use as many different and irregular dimensions of paper as possible? Or to see how small, messy, or light you could make pencil on paper? What secrets were concealed -- or revealed -- by over and over again scoring the paper with a sharp instrument and leaving behind absolutely no pigment at all?
Optimizing for Readability
With so little foolish consistency to work with, part of our initial challenge was to adequately scan such a wide array of varied documents. Being able to show the score marks equally well as bold, clear ink meant a lot of digital manipulation was needed. All documents were optimized for legibility: contrast and brightness were changed, a variety of filters were applied, saturation, hue and other adjustments were made. In some cases, this retouching seemed a nearly endless task: one particular image underwent nearly 10 rounds of retouching before we were satisfied. Happily -- given the realities of the project's budget limitations -- most documents were not as difficult to make legible.
The scanning and optimizing processes clearly brought out aspects of these documents which were only barely noticeable in the original. Due to increased size, contrast, clarity and legibility, it is easy to see marks on the digital images which might fail to catch the eye on the originals.
The scanning and optimizing processes also clearly brought out aspects of these documents which were simply NOT present in the original. Vivid coloration, moire patterns, and under-exposed darkened edges had to be digitally corrected wherever possible. And because each section of the large surveys might contain different types and combinations of pen, pencil or scoring marks, the levels of manipulation required for legibility varied throughout a single document. This left telltale and unavoidable color and shading variations which are apparent on several of the thumbnails.
During its course, this project was blessed by several happy surprises. First, the amazing leap in clarity of the enlarged and retouched documents delighted everyone who viewed them. Second, we feared that the thumbnail images when reduced to their required size would be tiny, unreadable blobs. Instead, they were crisp, clear, informative and pleasing to the eye. Third, we ended up with an unanticipated set of documents showing the largest of the surveys as single images at 200%+ magnification. These were an unexpected byproduct of constructing the thumbnail images. They allow viewers to study these surveys in detail with the integrity afforded a single image. Due to their size, they are not available on the website; they are on a CD-ROM and can be used at Special Collections.
A last surprise was a personal one. Spending hundreds of hours looking at Thoreau's handwriting, to me a few distinct styles emerged. There is the loose pencil scrawl of first draft documents which defines one end of the legibility spectrum. At the other end is the neat and formal inked hand found on final surveys, such as were given to his clients.
Along the middle, there are several other styles. One is a mostly-legible -- yet still rather loose -- style which occurs in both ink and pencil (above, right). Each time it was before me, I would start at the familiarity of it, but it took processing 50 surveys for me to finally place it. For confirmation, I emailed a snippet of it to my mother, asking her only to identify the author. She instantly replied, "Where did you find my father's handwriting? What does he mean here?" Pleasant memories of a beloved grandparent did hover throughout this project.
Both the staff of the Concord Free Public Library and Windfall Projects hope that the final product of this project will promote meaningful use of Thoreau's surveys at a level never
before possible. We encourage you to contact Leslie Wilson at LWilson@minlib.net with questions
or comments about the survey collection, and myself at info@WindfallProjects.com regarding
Many thanks go to Leslie Wilson and Robert Hall of the Concord Free Public Library, to Richard Stevenson and Barbara Peskin of Windfall Projects, and to AT&T for generous funding of this project.
Images of Henry David Thoreau surveys courtesy of Special Collections, Concord Free Public Library
Backgrounds: Uh, we're not sure where these are from. If you know, please contact us!
This website is a gift to the Concord community from ConcordMA.com, a full-service
Internet design and marketing company. 978 369-0113. PO Box 285, Concord, MA 01742 firstname.lastname@example.org