Double Fold

by Nicholson Baker

  At break time, I asked Dalton if I could take a look at the brittle book. It was called The Life Radiant by Lilian Whiting (Little, Brown, and Co., 1903). On the copyright page were blue and pale-purple stamps identifying it as the property of the New York Public Library, Astor, Lenox, and Tilden Foundations. On the inside cover, in red letters, was stamped:

  MICROFILMED

  Date Feb 07 1990

  TOO POOR TO BIND

  Every page I turned was legible; turning did no damage. On page 80, I read: “The demands of modern life absolutely require the development of some means of communication that shall obviate the necessity of the present laborious means of handwriting.” On page 82 was a quote from H. G. Wells on the discovery of the future. Pages 83 and 84 were gone, though—crumpled in workshops to demonstrate the urgent necessity of microfilm.

  I asked Dalton where he had gotten the book; he gave me an odd look. Perhaps it occurred to him that I wasn’t a librarian. “Uh, this is just something that when people go around here from the center to yard sales and things, if they see something like this they scoop it up,” he said. He smiled: “Always on the lookout!” The New York Public Library owns only microfilm now for The Life Radiant; its catalog record lists the 1990 reformatting as sponsored by a grant from the National Endowment for the Humanities. I suppose it is possible that the New York Public Library’s book happened to find its way to a yard sale in the environs of Andover, Massachusetts. It’s more likely that Dalton’s own staff at the Northeast Document Conservation Center microfilmed the book, and that the NYPL didn’t want it back afterward.

  At Dalton’s “School for Scanning” seminars (also brought to you by the Mellon Foundation and the NEH), librarians go to hear current thinking on embedded metadata, SGML tagging, lossless compression, TIFF formats, emulation, refreshment, and disbinding. Why disbinding? “To get a digital image that’s really well captured, it’s still best to disbind,” Dalton advised us.

  CHAPTER 32

  * * *

  A Figure We Did Not Collect

  How high, one wonders, was the national toll? How much destruction did we as a nation pay for? The National Endowment for the Humanities doesn’t know and doesn’t want to know. I asked George Farr whether the NEH ever kept track of the number of volumes that physically went back on the shelves after their contents entered the Paradiso of spools (though I didn’t put it that way), and whether the NEH required its grant recipients to provide any counts or percentages of physical books discarded or disbound as part of their project reports. “We have not done so”1 was Farr’s written reply. “NEH has never taken a position on the eventual disposition of brittle volumes that have been microfilmed to preserve their intellectual content. We believe that this is a decision that is more appropriate for the grantee libraries to make.”

  The most confidently flourished post-microfilm discard percentage that I have come across was offered by Carla Montori, head of preservation at the University of Michigan, in a posting to the Preservation Administrators Discussion Group. Montori writes:

  Analysis of 15 years2 of disposition decisions [at Michigan] indicates that something less than 40% of brittle originals are withdrawn after reformatting into or replacement by a stable format copy.

  Michigan has a long-established microfilming program, begun by Verner Clapp’s close friend Frederick Wagman. Perhaps, I thought, this forty-percent figure would make a good conservative estimate?

  I ran it by Ellen McCrady, editor of the Abbey Newsletter. She said she had no way of knowing, since libraries weren’t eager to make that sort of information public, but that it seemed on the low side. “I would guess that forty percent were saved whole,” she said. She told me to check the Association of Research Libraries’ annual preservation statistics, which might offer some clues. So I ordered a copy of the latest year then available, 1996–1997.

  You can learn all kinds of things from this staple-bound purple booklet.3 Harvard got $1.3 million in outside preservation money (a good chunk of it from the NEH); Boston Public Library got $3,522. Michigan digitized 1,350 “bound volumes” (they were no longer bound by the time they were digitized), and microfilmed 5,547. You can check how many volumes were deacidified, filmed, scanned, commercially bound, boxed, and treated to three different intensities of physical repair at each of over a hundred really big U.S. and Canadian libraries. Since 1988, according to these figures, which for technical reasons under-report actual levels, libraries microfilmed about one million volumes, most of the work paid for by the federal government. But you will not find in this purple book of preservation statistics—perhaps because the Association of Research Libraries is too tactful to demand it of its members—how many volumes were in fact preserved, in the old-fashioned sense.

  I asked Julia Blixrud, one of the people who compiled the numbers, whether there had ever been an attempt to track how many books survived reformatting. She said she didn’t think so. Was that possibly, I said, because it would be embarrassing to poll members for such a statistic? “That I don’t know, I’ve not been a preservation librarian,” Blixrud said. “I would guess that it was not as much of interest to us.” And she added, “The political nature of this survey has been to assist in some ways in getting funds to continue to do the preservation microfilming.” She told me to talk to Jutta Reed-Scott, who with Harvard’s Jan Merrill-Oldham designed the survey in the late eighties.

  “That unfortunately is a figure we did not collect,” said Reed-Scott. “I am not aware that anyone collected that. It would be a really difficult figure to keep track of. There are so many variables that go into returning items to the shelf.” (Oh, piffle. It would be no more difficult to track this figure than to track any other.) I asked Reed-Scott if the need for such a number had ever come up during the planning stage. “I do not recall that that ever arose as a question,” she said. I told her that my sense from talking to people was that there had been a sudden drop-off in preservational disbinds and discards around 1993. (The year, incidentally, of Thomas Tanselle’s book-burning piece.) “I think that’s exactly what happened,” Reed-Scott said. Now the problem was scanning. “That is the arena where the issue will become much more difficult than the preservation-microfilming arena,” she said. “You can film from a bound volume, with some difficulty, but it is certainly possible. It becomes far more difficult to scan from a bound volume, because obviously you get the distortion, if the margin is narrow.” In the evolution of scanning theory and praxis, we’re about where we were in 1953 in microfilm.

  There is a further fact to note about Jutta Reed-Scott. In the seventies, she was not a preservationist at all; she was a hard-core space freer-upper. In 1976, as Jutta Reed, then collections development librarian at MIT, she published a wonderfully cut-and-dried paper for Microform Review arguing that if you subscribed to the microfilm of the journal Daedalus, for example, and dumped all but current paper issues, you would save $1.35 per year in binding costs. Add to that the “dramatic reduction4 of storage costs of micropublications over hard copy periodicals,” as determined by the formulas5

  storage cost = ½av(t(t+1))

  and

  storage cost = ½bv(t(t+1))

  (where a is the yearly cost of storing one bound volume, and b is the yearly cost of storing one volume copied onto microfilm; where t is the storage time in years, and v is the number of volumes per year), and you really begin to save, save, save. Assuming an annual storage cost of $0.38 per journal volume and $0.034 for microfilm, our soi-disant preservationist calculates that if you replaced bound volumes of Scientific American with microcopies, you would over a twenty-year period save over $1456 in storage costs. (Of course, you will no longer be able to interpret Scientific American’s color-keyed illustrations properly, and over that same period you’ll have paid University Microfilms a fortune for the microfilm subscription.) “In the long run microforms will increase the storage capacity of present library buildings and can postpone the construction of additional
storage space,” Jutta Reed unwaveringly writes.

  Perhaps it isn’t so surprising, then, that Reed-Scott exhibited an indifference, a listless lack of curiosity, as to how many original volumes were saved after they were microfilmed, since in an earlier life she believed in the “decisive economic advantage” of the dump-and-replace method without any consideration of embrittlement whatsoever.

  CHAPTER 33

  * * *

  Leaf Masters

  The NEH didn’t know and didn’t care; the national preservation statistics were mute on the subject; but still I kept an eye out for estimates and demi-disclosures, because I believe we need to know how much was lost. In 1984, David Stam, then director of the New York Public Library, wrote that “a heavy proportion1 of our embrittled material has been discarded after microfilming” (by “our” he meant in the United States), and he said that “the practice of disposal has often included materials that could have been adequately preserved in the original.” There is, however, no way to quantify the words “heavy proportion”—does it mean over half? I called Yale’s Paul Conway back and asked him whether a fifty-percent discard figure was a reasonable guess through 1993 or so. He said that “fifty percent rings true” for the years between 1983 to 1993. “But since then libraries are trying a lot harder.”

  Gary Frost, who was for years in the business of making sumptuous preservation photocopies (sometimes with color-copied title pages) at a place in Texas called Booklab, wrote in 1998:

  If the current compulsion toward digital scanning of library materials is really the dawning of a New Age era of microfilming what is the preservation implication? The implication may be the same; that the new technology will be considered a preservation method that results in the discard of library materials. The microfilming process, by one estimate, has resulted in the discard of over 60% of the originals recorded.

  Over sixty percent? Was that a reasonable estimate? Gary Frost is an interestingly ambivalent—one almost wants to say “tortured”—soul. He strongly believes that librarians should continue to function as custodians of the “source original.” Conservators should “support developments in storage facilities and delivery systems that enable the survival of originals.” He told me that Michael Lesk’s and Brother Lemberg’s idea of discarding millions of books in favor of networked digital page-scans is “a right-wing paramilitary objective—or no, a left-wing paramilitary objective.” He regrets the “very strong undercurrent in the USA toward disposability, toward favoring clean copy over soiled original.”

  But in producing Booklab’s fabled preservation photocopies, Frost’s team of operators severed spines just as microfilmers would, to get pages to lie flat. Frost gave his mutilés de guerre a nice name, though: he called them leaf masters. Libraries were supposed to put their pristine, acid-free Booklab photocopies on the shelf and store the soiled leaf masters in cool sanctums, using them like the frozen sperm from great racehorses whenever the circulating photocopy had done too many laps. Originals were not for “eye reading,” but for “machine reading”—for the infinite propagation of copies. (“I’ve never seen a book I couldn’t copy,” Frost said, when I asked about the extent of crumblement or dusthood.) Not so many libraries were interested in storing boxed or shrink-wrapped bundles of loose pages, though; the leaf masters often went in the compost pile.

  “We’ve cut two or three hundred thousand books here in flat platen work,” he told me. “It is what it is. I know that the types of materials that we were working from were such combat-battered brittle materials anyway that there wasn’t any circulation left in them. I also realize the dilemma of in any way projecting a preservation agenda based on destruction, even though that’s been possible in a couple of weird ways.”

  I wondered what percentage of the disbound Booklab books were kept by libraries afterward. If I was to go to a university and ask to see their disbound, shrink-wrapped Booklab originals, I said, what would my chances be of seeing something?

  “I would say that your chances are probably very minimal,” Frost answered. “Certainly in the ten- to twenty-percent category of stuff that we did. I could be wrong.” The libraries at the Metropolitan Museum of Art don’t throw away their leaf masters, according to Frost. “They don’t believe in brittle books. That’s almost seven or eight or nine libraries right there, in that one museum. But that’s kind of the exception that proves the rule that most of these would be thrown away.”

  I asked him who had been the verbal source original for the estimate that microfilming had resulted in the discard of “over sixty percent” of the books recorded. John Dean, of Cornell, had come up with that a long time ago, Frost said.

  “I don’t think so,” said John Dean, when I reached him. “I have no basis for that kind of number at all.” Dean did confirm a discussion in 1993 (at an annual meeting of the NEH’s Office of Preservation) of how much to keep after microfilming. “Faculty and bibliographers and curators were saying, ‘Hey, if you’re going to throw this stuff away, I don’t want it filming.’ ” The idea that libraries should now digitize their goods so that other libraries can throw out their copies is one he attributes to “the brutalist school.”

  The oddity here is that Cornell has been a hotbed, or testbed, of crypto-brutalists—but that’s a matter for a later chapter. Perhaps the sixty-percent estimate Gary Frost had in mind came not from Cornell’s John Dean, who disowned it, but from Yale’s Gay Walker. Walker wrote in 1989: “Based on a non-scientific survey2 of the field, the majority of filmed volumes are subsequently withdrawn from collections.” The survey to which Walker refers was eventually distributed as part of an informational compilation called Brittle Books Programs, produced by the Association of Research Libraries. “Of all responding libraries,”3 write the surveyors, Jan Merrill-Oldham and Gay Walker, “nearly half discard 90 percent or more of the original copies of volumes after they have been reproduced.” Almost a quarter of the thirty-five libraries surveyed kept none of the volumes after reformatting, they found; two libraries declined to answer. Walker does not volunteer Yale’s own percentage; but included in this compilation is another 1986 document in which she writes that items at Yale designated for withdrawal should “have all ownership marks removed4 or marked out and a ‘discard’ stamp applied to the inside front flyleaf and the inside back cover as well as the title page. . . items are sent to the Gifts and Exchanges Unit when a truckload has accumulated.” In a 1987 article for Restaurator, Walker offers another hint: “In the great majority of cases,5 the very deteriorated, brittle, disbound original is withdrawn from the collection and often placed in a library booksale for a minimal amount.” Walker was “Arts of the Book Librarian” at Yale when she wrote that.

  Walker and Merrill-Oldham’s survey is, in any event, the best estimate I’ve found of discarding practices at the height of the brittle-books agitation. To be cautious, though, let’s say that only half of the books, not (as Walker says) “the majority,” were thrown out. Between 1988 and 1993, the NEH paid for the microfilming of about 500,000 books, and the Library of Congress’s Preservation Microfilming Office filmed another 150,0006 volumes or so. At a fifty-percent retention rate (which is, for the Library of Congress of that era, extremely conservative), 325,000 books were removed from U.S. libraries as a direct result of federal money. Add the newspapers to that.

  But the losses exceed that number. In the same paper, Walker wrote that the “filmed copy must be a perfect one—other copies of the book will be discarded upon the strength of the listing.” Does this still seem strange—that libraries would cast aside their own copies of books simply because they have been judged by others worthy of filmed preservation? It was never a strange idea within the preservation movement. Recall that Robert Hayes had built it into his cost-benefit analysis: the creation of a film would have a multiplier effect around the country, triggering local bibliectomies. (Sometimes a library would buy the microfilm from the filming library to replace its original; but sometim
es the library would simply feel better about deselecting its original in the knowledge that preservation-quality microfilm was potentially there to be bought, even if it wasn’t actually bought.) Hayes figured that five physical copies would disappear from libraries for every book filmed, but let’s say, more conservatively, that from 1988 through 1993 fifty percent of the filmed books spawned three physical disappearances—one at the filming library and two of duplicates somewhere else. And let’s say that none of the remaining fifty percent (i.e., those books that were physically reshelved post-microfilming) prompted any parallel discards elsewhere—highly unlikely, but also conservative. That triples the loss-estimate and brings it to 975,000 books. Almost all these books were old and out of print, so the replacement cost, assuming that these original editions could be found on the used-book market, and ignoring costs of recataloging and reprocessing, is high—say (again conservatively) forty dollars a book. As a very rough, lowball guess, thirty-nine million dollars’ worth of originals left our nation’s libraries, thanks to federal largesse. It’s as if the National Park Service felled vast wild tracts of pointed firs and replaced them with plastic Christmas trees.