Friday, May 31, 2013

First Editions of Whitman's Leaves of Grass


Following on last week's post which mapped books based on some longer-term research, I wanted to show what kinds of simple geographic visualizations are possible using data that's already extant and well-curated.

Given that today is Walt Whitman's 194th birthday I thought I'd post something related to the Walt Whitman Archive which is one of the longest-lived and most comprehensive digital projects out there.  Among its many virtues is the willingness to give users access to raw data as well as interpretive content. I have a soft spot for comparative bibliography and was really impressed by Ed Folsom's multi-year study of all 158  known copies of the 1855 first edition of the Leaves of Grass. The survey, published in the Walt Whitman quarterly in 2006, came with a downloadable data set recording bibliographic details on all known copies. The 1855 Leaves has a complex publishing history and many of the copies differ from each other. Perhaps the most remarked-upon difference between the copies, which Folsom's survey highlights, are the many states of the infamous portrait of Whitman which appears on the frontispiece - particularly, the size of his crotch.

So-called "flat" (left) and "bulging" (right) versions of the 1855 frontis. Figure 5 in Folsom's magisterial Whitman Making Books/Books Making Whitman: A Catalog and Commentary (Iowa, 2005)

The thoroughness of the survey can be seen in the detailed comments on the state of these frontispieces, for example this one on Library of Congress copy shelfmark PS3201 1855a c.1 - which might also be one of my favorite bibliographic descriptions ever:

"The crotch appears to be a version between the flat and enhanced crotch. There is considerably more shading to the left of the small bulge, but it is not yet fully enlarged. The bottom of the pants is differently engraved than any known version"

It's one thing to see all of this great detail in spreadsheet form but I thought perhaps Folsom or other Whitman scholars might be interested in seeing a direct visualization of the geographic dispersal of the physical artifacts of Whitman's work. Below I've mapped the known copies by their holding location showing their concentration by institution or private owner. The majority never left the original 13 states on the east coast.



For a full look at the visualization and data see http://geocommons.com/maps/169120

Tuesday, May 21, 2013

The Penn 100: Unique Early English Printed Material in a Research Library Collection


ESTC R179291 (Only known copy: Penn EB65.A100.689K)

In the fall of 2011 I asked the administrators of the English Short Title Catalog (ESTC) if they could provide me with a list of all the records in their database which reported holdings only at the University of Pennsylvania –  i.e. “unique to Penn” imprints. The ESTC has as its purview all printed material produced in the British Empire or in the English language from the dawn of printing to 1800.[1] The ESTC sent me back a list of 313 record numbers which they identified as having a holdings location only at the University of Pennsylvania. On and off over the intervening time, I’ve gone through the list, matching record numbers with titles and shelf locations and examining some of the records more closely.

Though ESTC imprint data has been collected and corrected by experts over many decades, I was well aware of the problems and possible errors inherent in such a dataset. To that end I looked carefully at the records and, in many cases, the physical books themselves to determine which were likely to be actually unique holdings.

Of the 313 imprints listed by the ESTC as existing only at Penn, I found clear evidence that 51 were also held by other libraries. A variety of factors explain these errors, some of which are instructive for users of the ESTC at large. In the majority of the cases simple cataloging aberrations caused mis-reporting.[2] Others were so-called “Wing Ghosts” i.e. listings in D.G. Wing’s bibliography that could not be identified in retrospect.[3]  In some cases, Penn had cataloged volume parts of a series individually creating “unique” records where entries for the series as a whole existed elsewhere.[4]  In other cases, as with the 1783 false London imprint Memorie sulla Bastiglia del celebre Sig. Linguet scritte, libraries outside the ESTC’s sweep proved to have relevant holdings.[5]

Examining the list of 262 remaining “unique” imprints with an eye towards understanding their survival in only one known institutional location is a larger project but I thought I would introduce the findings in aggregate first.

Chronology:

The earliest “unique” title found was printed in 1555 – a London second edition of a translation of Erasmus. This is the only 16th century imprint on the list and has unfortunately been missing from the library for some time.[6]

Taking the other 262 imprints a clear chronological theme emerges, representative of larger trends in book production. Of these, 26 date from the 17th century and 236 from the 18th century.
 
Chart showing number of "unique" ESTC imprints held by Penn grouped by five year intervals

Above you can see the near exponential increase of “unique” holdings by year of publication. Accordingly, the 5-year period 1795-1800 claims the highest number of imprints on the “unique” list with 29.

Geography:

As one might expect London dominates the list by printing location. Given Penn’s collecting history however, there are also a number of western hemisphere imprints represented. One of my early finds from the list was a unique 1719 Jamaica imprint, which is the oldest example of Caribbean book printing extant in the Americas. Below is a list of the 262 "unique" imprints by place of printing. Detailed map here.


Europe: 

London: 155.
Elsewhere in England: 13. This includes two from Wolverhampton and one each from Reading, Rochester, Shrewsbury, Southampton, Tunbridge, Manchester, Newcastle, Lincoln, Eaton, Canterbury, and Bristol.

Ireland: 30 from Dublin, 1 from Armagh, and 1 from Belfast

Scotland: 6 from Edinburgh and 5 from Glasgow

The Continent: 3. One each from Amsterdam, the Hague, and Mainz as well as one false Florence imprint (likely London).
Colophon from ESTC N67272
Americas:
Caribbean: 3. Two from Kingston, Jamaica and one from Antigua.

North America: 43
Philadelphia: 31, Boston: 3, as well as nine from elsewhere in the 13 colonies including one each from Lancaster, Pa.,Whitehall, Pa.,York, Pa., Peacham, Vt, Hudson, Ny., Hagerstown, Md., Exeter, N.H., Baltimore, Md., and Annapolis, Md.

Availability:

While ECCO, EEBO, and Digital Evans contain a wealth of digitized content, only a small fraction (a little over 8%) of the items on the Penn "unique" list are available in online databases. This number would of course change significantly if one were interested in any edition of a text rather than the specific imprint held here.

Collections data:

Items like those identified here as “unique” to Penn don’t come around every day. In fact, a large number of these “unique” items come from collections carefully developed over many years by individuals and institutions around particular themes. The ESTC data do a nice job reinforcing the importance of focused and deep collecting. Within the data, the collection that really shines through is Penn’s excellent Singer-Mendenhall collection of English fiction to 1820. A whopping 32 of the imprints, all novels, come from this collection, began in the 1920s by Godfrey Singer, a Penn student. Other focused collections represented in the "unique" list are the Horace Howard Furness Collection, with nine imprints, nine also from the Curtis Publishing Company Collection of Franklin Imprints, seven from the Teerink Swift collection.[7], four from the Edwin Forrest Library Collection, three from the Yarnall Library and three held at the Biddle Law Library. Interestingly, an additional four “unique” items come from an artificial collection we call “Founders” indicating books that were present in the University of Pennsylvania library before 1829. Contrary to the focused collecting of rare material made by later collectors, the four “unique” books imprints in this collection were purchased for mundane use by the students of the early University and include a (likely pirated) 1762 Dublin reprint of Hume's multi-volume history of England. These texts, which were likely readily available at the time have not survived elsewhere likely thanks to years of library weeding and the purchasing of the latest available editions.

The Penn 100:

At the very bottom of the page I have provided a link to the spreadsheet containing all 262 “unique” items I identified, but I wanted to go further to locate those works whose integral text exists nowhere but Penn. After filtering out translations where foreign language versions exist, books with other extant editions, and any sort of reprinting, I was left with just 100 items and a list spanning 1659-1800, only 14 of which have digital surrogates. These truly unique texts might be some of the most fruitful to examine in detail in the future:


What might we say about survival and the circulation of print publication based on this group? Surprisingly only 13 of these 100 are single sheet broadsides. Given how ephemeral these publications are I expected them to dominate the list of wholly unique items. Likewise, only four of the 100 are related directly to the University of Pennsylvania. Instead, many of the publications on the list are substantial works. Foremost amongst these, and the clear leader on the list, are the 23 novels dating from 1765 to 1797 which do not appear in any known institutional collection. Take for instance the 1787 novel Lord Winworth; or, the memoirs of an heir, the copy surviving at Penn is noted as the second edition but though an advertisement survives announcing the publication of the first edition, no copy seems to exist. Further proof that distribution and reprinting of texts doesn’t necessarily lead to their survival. Going forward, I’m eager to see the Early Novels Database project and its students shed some more light on these rare novels.

Also of interest on the list are the 10 imprints which are textbooks, guides, or conduct manuals. These include such scintillating titles as “An essay on a method of finding the solid contents of packages, by an easy addition of three numbers only.” Textbooks and school manuals were commonly used by the hundreds but due to the nature of their use, relatively few have survived to this day. Not all of these 100, though printed in British-controlled areas, were printed in English. In addition to a few Pennsylvania German imprints, perhaps my favorite in this category is the unique 1795 Le Chansonnier républicain a book of French revolutionary songs printed in Philadelphia for the émigré community.

Any of these “unique” works would make an excellent research project for a curious student or scholar and I hope providing these lists helps launch a conversation about the nature of library collecting and the place of bibliographic data in print history.

[For the complete list of all 313 imprints listed as "unique" to Penn by the ESTC with my notes and comments see CSV file here (filter by the field "Likely Unique" to see the final 262)]

****
I’d like to thank Ginger Schilling and Brian Geiger at the ESTC/UC-Riverside for being incredibly helpful in providing the data for this survey. Also note that the data in the ESTC is constantly being changed and updated. I would love to know of any additions or subtractions to these lists as appropriate – I’m sure it’s not the final word! 



[1] For more on the ESTC and plans for the future see their blog at: http://estc21.wordpress.com/
[2] See the example of  N31145 which was perhaps not caught because of a slight misspelling in the title. This imprint seems to actually be T132866 held by several libraries.
[3] For an example see R180866 which is actually held at Penn as R231132.
[4] For instance Penn’s volume of the Merry Wives of Windsor from 1734 [N475057] which is actually just one volume from the larger set of Shakespeare’s works held by many libraries as T54097.
[5] I was able to locate a copy at the University of Mannheim which apparently has never sent in its holdings to the ESTC:   Broadly speaking, the ESTC is only as good as the records which libraries chose to submit to it. For an example closer to home see W20303  Which mentions a copy in the College of Physicians of Philadelphia in the notes field but includes no holdings for that institution in the list of libraries.
[6] This Erasmus volume is ESTC S92484. It has been missing since at least 1974. See John Fleischauer, "A New Sixteenth-Century Translation of Erasmus," Papers of the Bibliographical Society of America, 68 (1974), pp. 164-6. [7] For more on this collection of material relating to Jonathan Swift see Daniel Traister, "The History of the Herman Teerink Collection of Jonathan Swift at the University of Pennsylvania Library," Swift Studies, 10 (1995), 80-88.