Thomas Jefferson's Libraries:
An Annotated Bibliographic Database

• Important Characteristics of Certain Manuscripts
• Judging Jefferson's Ownership
• Editorial Method
• Editorial Decisions
• Choice of Edition in the Thomas Jefferson Portal
• Project Personnel
  

Introduction

Thomas Jefferson preserved voluminous records of his intellectual interests. These take many forms: catalogs of his various libraries, lists of books to acquire, recommended reading lists, notes in Commonplace Books, and a multitude of comments in his prolific correspondence. As a result, the development of Jefferson's library collections can be tracked to an extraordinary degree. The record provides a wealth of information on what he read, what he valued, what influenced him, and what contributed to the foundations of his thinking. Moreover, Jefferson's intellectual standing and his wide sphere of influence makes this data more broadly relevant and useful in studying a wide cultural and political milieu.

The Thomas Jefferson's Libraries project will compile, for the first time, information about the books that Jefferson is known to have owned, or expressed an interest in, during his lifetime. Information from seven lists and catalogs will be edited, matched to bibliographic information, and made available worldwide through the Thomas Jefferson Portal (TJPortal) online catalog. The Machine Readable Cataloging (MARC) records will be expanded with notes, subject terms, series links, and other relevant details to enhance retrieval options. Transcriptions of Jefferson's notations about titles and editorial commentary about Jefferson's ownership of, and interest in, each title will augment the bibliographic information. To enhance searching retrieval and to expand the possibilities for making correlations among titles, additional data will be provided: notation of the library or libraries in which the title was contained and the subject category into which Jefferson classified each title.

Pilot: "Geography--America"

To demonstrate the scope, search capabilities, and potential for use by students and scholars, a pilot project covering some 350 titles classified by Jefferson in his "Geography--America" section is now available through the TJPortal. This pilot provided an opportunity to establish standards and practices that will be applied throughout the project. Bibliographic and editorial procedures were defined to assure consistent and authoritative data. The Preface provides background information on how procedures have been established and the bases for key editorial decisions that users should keep in mind when interpreting and using the data.

Preface

Compilation of this database entailed a number of challenges, and the examples and explanatory notes included below illustrate some of the complexities and the rationale behind the decisions that have been made. The starting point for data compilation -- short title notations in manuscript lists -- has dictated that special editorial and bibliographic methods be implemented. Every effort has been made to transcribe the manuscripts in their entirety to strengthen the case for verifying titles, editions, and Jefferson's ownership.

Jefferson's Copy of Homer's Iliad. (Click on Image to Englarge)

The transcriptions compiled from lists and catalogs represent the best efforts to preserve accurately the original record. In accordance with current documentary editing practice, records in Jefferson's own hand are included in their entirety, as even crossed out or erased entries provide indications of his interests. The editorial apparatus indicates where erasures, cross-outs, and illegibility have compromised transcription. Incompletely erased works and partially obscured words are included in the interest of completeness and accuracy. In some cases, entries are included based on best editorial judgment with the understanding that experts from around the world will now have access to this information and may provide comments and corrections. The project provides a quick link for users wishing to supply feedback.

Presence of a title in one of the lists does not necessarily confirm Jefferson's ownership, though it does indicate his interest. In many cases, his ownership is corroborated by internal evidence, other records, or mention in correspondence, and such evidence is provided where known.
Detailed analysis of ownership questions is provided below.

In cases of multiple works bound together, Jefferson often, but not always, listed the individual titles. E. Millicent Sowerby's examination of the works purchased by Congress in 1815 (in her Catalogue of the Library of Thomas Jefferson, 5 v., 1952-1959) led to identification of works that can be appended to those listed in the manuscripts. These works are individually cited in the database with notation of the bound-together title under which they were included. Researchers who wish to view the original documents may consult the "List of Libraries Documented" section on this web site for information on repository and facsimile availability.

Titles and imprints from Jefferson's notations and/or Sowerby's annotations are matched, as far as possible, with bibliographic information from authoritative databases, including the Library of Congress catalog and OCLC WorldCat. This provides standard forms of names, titles, subjects, and edition information for Jefferson titles based, in most cases, on an existing copy of the title in a library or repository. Where edition information is unknown or obscure, a bibliographic record is utilized that best matches that title. For details on the decision criteria used, please refer to "Choice of edition in the Thomas Jefferson Portal."

Important Characteristics of Certain Manuscripts

Great Library

Jefferson's Boyhood Copy of Virgil's Aeneid. (Click on Image to Englarge)

Jefferson began to compile this list at some point in the 1770's - probably shortly after the fire at Shadwell in February 1770 that destroyed his first book collection. This manuscript provides an abundance of internal evidence as a register of his wide-ranging interests in printed works. He continued to use this list for four decades as a record of titles that he owned and, in some cases, those he intended to acquire. In the Thomas Jefferson's Libraries project, "Great Library" is used to designate this list, the original manuscript of which is held at the Massachusetts Historical Society (MHi). The Thomas Jefferson Papers web site presented by MHi, lists this manuscript as the "1783 Catalog of Books." The notation on page five is particularly valuable, because it allows us to determine which titles he owned as of March 6, 1783. This note begins with a check mark in the left margin, followed by, "this mark denotes the books I have. those unmarked I mean to procure. 1783 Mar. 6 / 2640 vols." The evidence provided by Jefferson's check marks in the left margin is the primary basis for reconstructing his library at this particular date, and these 2,640 works, a discernible subset of the Great Library, have been designated as the "1783 List" in this project.

Jefferson's methodical attention to organizing his books is readily apparent in this document. He used a subject classification scheme, derived from Francis Bacon's categories of human knowledge, to separate the Great Library into 46 "chapters," with further division of many chapters into subsections. For instance, chapters on history were separated into geographical categories. The complete list of chapters and sub-divisions is provided at the Massachusetts Historical Society's Thomas Jefferson Papers web site. Jefferson left space on several pages for additional acquisitions within the categories, but in some instances these attempts to allow for expansion were inadequate. The inevitable result was a plethora of interlineations as he attempted to maintain the rational categorization of his library titles.

1789 List
This list seems to indicate reliably the titles purchased during Jefferson's five years as ambassador to France. This list may have served as a packing list of titles for shipment back to the United States. There are diagonal cross-outs of every work, probably a system to mark titles as verified for shipment. The vast majority of titles are also listed in the Great Library and in the Library sold to Congress in 1815. Sowerby links some of these works to payment notations in Jefferson's Commonplace Books, and most of the works in this manuscript include a notation of the price for the work. In addition to the consistent diagonal cross-outs, there are some horizontal cross-outs that extend across the entire listing. Most works that are crossed out horizontally are found later in the list, and these may have resulted when a work was sent to a contact in America and subsequently replaced. Other cross-outs that are not re-listed may have indicated the absence of these works during packing for shipment.

Retirement Library
The Retirement Library manuscript, spanning eleven years, contains relatively few revisions. There are gaps in some places, but proportionately there are far fewer cross-outs and interlineations than in the Great Library manuscript. The familiar chapter classification is used, but with under 1,000 titles listed, there is less use of subgroups within the chapters. One explanation for the comparative orderliness of the Retirement Library manuscript may be that Jefferson's reading habits during his retirement years were well established. His interest in contemporary politics and culture did not wane, and he was able to spend considerable time with classical works, especially Cicero (compare his correspondence: TJ to John Adams, 5 July 1814; TJ to John Adams, 10 December 1819; TJ to Thomas W. White, 8 December 1824). There is less evidence of fruitless seeking after certain titles as evidenced by the lack of cross-outs and re-listings.

Judging Jefferson's Ownership

The following guidelines and examples will prove useful in explaining how Jefferson's ownership of a work is determined. The Sowerby and Trist lists are reliable indicators of what was sold to Congress in 1815, thereby providing confirmation of his ownership. Other reliable corroborators include internal evidence in the 1789 List and evidence in his correspondence and Commonplace Books. Ownership of the works listed in the Retirement manuscript is confirmed by nearly identical listings in the 1829 catalog of the Poor auction at which Jefferson's Retirement Library was dispersed. On the other hand, some listings in the Great Library manuscript may never have been obtained. Jefferson's correspondence with numerous booksellers may well have been the source of detailed descriptions of works in which he had interest but was unable to acquire.

An example of an entry that Jefferson listed but probably never procured is found on page 168 of the Great Library manuscript, lines 25-26. His detailed description of this work, "Lederer's discoveries in Virginia & Carolina in 1669. & 1670. by Sr. Wm. Talbot. London printed for Saml. Heyrick. 1672. 4to." was partially rubbed out. It does not reappear elsewhere in the chapter, does not bear a check mark, and was not included in any other list. Thus, it is probable that Jefferson never procured this work.

An example of a work in the Great Library manuscript that Jefferson definitely owned is "Strabo. Gr. Lat. Casauboni fol." listed on page 158, line 5. There are numerous indications of ownership including a check mark with a period, larger script with wider line spacing, and inclusion at the beginning of the chapter. In addition, the work was listed in Trist and Sowerby. The textual evidence indicates that Jefferson most likely owned the work in 1783. Jefferson's fondness for the work is indicated by the purchase of a second copy for his Retirement Library, listed as 310 in the Library of Congress manuscript, and corroborated in the Poor auction catalog.

Duplicate listings in the Great Library manuscript raise questions that can be illustrated with three entries on page 168, lines 1, 2, and 9. None of these are erased or crossed out, and yet each title appears elsewhere in the manuscript with variations in the bibliographic information. The size of the script, line spacing, and position on page 168 indicate that the listings were made prior to 1783, and the absence of check marks indicates that he did not own copies as of that date. Thus, the internal evidence suggests that the listings were notations of interest and intent to acquire, and that he neglected to cross out these initial listings after he procured them. Jefferson's failure to erase or cross out the earlier listing of works he later procured was the exception and not the rule.

Page 167 contains what appear to be later listings of the three titles from page 168. The work listed on line 1 of page 168, "Voiage au pays des Hurons par Gabriel Sabard Theodat. Paris. 1632." misspells the name of the author. Jefferson corrected this misspelling from the edition procured and listed on page 167, line 22, "Voiage au pays des Hurons par Sagärd. 2.v. in 1. 12mo." The work listed on line 2 of page 168, "Moeurs des Sauvages de l'Amerique par Lafitau" was relatively generic in its description, and the entry on page 167, line 26, "Lafitau. Moeurs des Saüvages Americains. 2.v. 4to. Par. 1724." is more specific, as if recorded from a procured copy. The work entered on line 9 of page 168, "Histoire de la Nouvelle France avec les fastes Chronologiquis du nouveau monde par le pere Charlevoix. 3.vols. 4to." was quite specific in its long title, number of volumes, and size. It would seem, however, that Jefferson was able to procure a different edition based on variations in size and number of volumes in the entry on page 167, line 36, "Histoire de la Nouvelle France par de Charlevoix. 6.v. 12mo."

Another listing on page 168 of the Great Library manuscript provides a variation on the question of ownership. The work listed on line 22, "Discourses of Virginia. p. 4to." bears the check mark and period, but is not found in any other list of Jefferson's books. As the check mark, size of script, line spacing, and location all indicate Jefferson's ownership as of 1783, the absence of that work from either the Trist or Sowerby lists indicates it may have been lost sometime after 1783. It is possible that Jefferson held this work back from the sale to Congress, but it does not appear in either of the lists that reference Jefferson's post-1815 libraries (Poplar Forest and Retirement Library). An explanation based on the evidence is that this item was owned as of 1783 and subsequently lost.

Editorial Method

Compilation of this database begins with transcription of a chapter in the Great Library manuscript and the same chapter in the 1789 List. All transcriptions are proofread and assigned a unique record number in the database. In the case of multiple editions of the same work, these are each set up as unique records. When a work appears in multiple manuscripts, the information from every manuscript is added to the database record. Listings that are difficult to read accurately from facsimiles or digitized images of the manuscripts are noted for future examination in the original manuscript and consultation with the editorial team.

Once all the works in a particular chapter or subject category have been transcribed, cross-referenced, and compiled, the data is converted into MARC format. Titles are then searched in established bibliographic databases to arrive at the best match. This process, discussed in greater detail below, is challenging because it entails determining which of the many eighteenth and nineteenth century editions was the work recorded by Jefferson.

It should be noted that Sowerby included only a part of the Great Library catalogue, omitting anything that seemed to be crossed out or erased, even in cases where other evidence indicates that Jefferson owned the work. Her primary goal was to determine exactly which works had been sold to Congress in 1815, and therefore, her catalogue is not an inclusive record of all the works represented in the Great Library manuscript. In contrast, this database seeks to record every title that can be linked to Thomas Jefferson.

Editorial Decisions

Particular challenges were faced in the process of transcribing and interpreting the handwritten information, especially from the Great Library manuscript. Intricate interlineations along with erasures and cross-outs present obstacles in the quest to include everything found in the manuscript. This is exemplified on page 135 in the Politics chapter, where legibility is impaired and complete accuracy in transcription is called into question. Whenever possible, obscured or partially erased works have been included. An example is, "Coloniae Anglicanae illustratae. 4to. Baker. 1763." on page 149 listed on a separate line under the work "Hazard's Historical collection of American State papers."

All too often there is insufficient evidence upon which to make a judgment as to the original listing. Illustrative of many instances are the first two works listed on page 158 at the start of Chapter 30 . Both works were substantially erased and two new titles overwritten in black ink, in contrast to the brown ink seen elsewhere on the page. The only safe editorial conclusion is to exclude the original two entries. Including such highly abbreviated works in the database is too dubious to serve any purpose, and the hope is that subsequent disclosure through technical, editorial, or corroborative means will reveal unknown titles.

Choice of Edition in the Thomas Jefferson Portal

To arrive at the bibliographic information to be displayed for each title in the Thomas Jefferson Portal, known imprint information from Jefferson's notations and/or from Sowerby's annotations were matched as closely as possible with data from established bibliographic databases, including the Library of Congress catalog and the OCLC WorldCat database. This provides a full range of information for every title -- author(s), subjects, content, edition, notes, etc. -- based on one or more existing copy of the title in a library.

In many cases, a choice had to be made between a number of similar bibliographic records found in these databases. Firstly, title and author information had to match information from Jefferson's listings and/or Sowerby's entry. For some titles, catalogs and booklists of library collections contemporary to Jefferson^[1] were consulted to assist in this identification process. Next, every effort was made to match publication dates and physical sizes recorded in Jefferson's notations and/or identified by Sowerby to arrive at the most likely edition of the title for inclusion in the TJPortal. Where more than one imprint exists with the correct title, author, date, and other imprint information, preference was given to the record found in the Library of Congress catalog for a book in the Rare Book Collection, or the record that describes a known Jefferson copy of the title. Future steps in the Thomas Jefferson's Libraries project will include a more comprehensive effort to identify extant copies of Jefferson's books and to document in the TJPortal their current repository locations. Updates to the bibliographic information will also be made, where necessary.

When it was not possible to determine the actual edition of a title from Jefferson's listings or Sowerby's annotations, the record was utilized that best matched other known information, such as the date of the manuscript or catalog in which the title was listed, the period in which Jefferson is believed to have owned the title, or mention of the title in his correspondence or other contemporary catalogs and booklists. Again, preference was given to the record found in the Library of Congress Rare Book Collection. In some cases, Sowerby stated that for a particular title no such edition as described by Jefferson has ever been found or is known to have ever existed. It is important for users to note that while the title and author data in the TJPortal are accurate, accompanying edition information such as date, publisher, or size of the imprint may not match the specific items that Jefferson owned or that he had in mind to procure when he listed a title.

When no bibliographic record could be found that matched Jefferson's listings, or when his short title was insufficient to allow determination of the actual title, the transcription of his notation has been entered, within square brackets, [ . ] , as the title of the work in the Thomas Jefferson Portal.

Project Personnel

John Barden, Editorial and Bibliographic Consultant, Richmond, Va.

Thomas Baughn, Editor and Transcriber, Silver Spring, MD

Donohue Group, Inc., Data Mapping and Technical Processing, Wethersfield, CT

Jack Robertson, Foundation Librarian, Thomas Jefferson Foundation

Endrina Tay, Coordinator of Cataloging and Metadata, Thomas Jefferson Foundation

Carrie Taylor, Collections Manager, Thomas Jefferson Foundation

Douglas L. Wilson, Editorial Advisor, Galesburg, IL

Chad Wollerton, Monticello Webmaster, Thomas Jefferson Foundation

« Back to Jefferson Library Home

Footnote

[1] Sources include:

Rutland, Robert A. "Well Acquainted with Books": The Founding Framers of 1787, with James Madison's List of Books for Congress. Washington : Library of Congress, 1987.

Wolf, Edwin. The Library of James Logan of Philadelphia, 1641-1751. Philadelphia : The Library Company of Philadelphia, 1974.

A Catalogue of Books Belonging to the Library Company of Philadelphia : A Facsimile of the Edition of 1741, Printed by Benjamin Franklin, with an Introduction by Edwin Wolf. Philadelphia : Library Company of Philadelphia, 1956.

Hayes, Kevin J. The Library of William Byrd of Westover. Madison, Wisconsin : Madison House, 1997.