None of that would matter much, except to paper scientists, if the fold test, allowing for all of its invalidating irreproducibilities, were a useful rough indicator of paper’s ability to do what readers ask of it. Is it? We ask of a book that its pages remain attached to their binding and turn. Maps must fold and unfold as a condition of use, dollar bills must survive pocket-crumpling and repeated wallet-bound contortions — book paper must turn without breaking. The late Klaus Hendriks, a scientist at the Canadian Conservation Institute, wrote:
While folding endurance
4
is more sensitive to changes in paper than any other strength test, papermakers essentially use it only in the manufacture of paper for applications such as bank notes and maps.
The fold test, in other words, is the wrong test to be using on books. Indeed, Hendriks rejects other currently available mechanical measurements as welclass="underline" “None of the commonly used paper tests5 bear any resemblance to the way a paper document or book is handled in practice,” he writes.
Nonetheless, Barrow favored the fold test above all others because, he contended, it “simulates the bending of a leaf6 to and fro in a book in use,” and because it “seems to lend itself most readily to analysis”—meaning it made the best graphs. (In one of the Barrow Laboratory’s books,7 Verner Clapp and Barrow are photographed together as they admire a large wall-mounted graph of the precipitous decline in paper’s fold endurance; the inverse, in a way, of Fremont Rider’s exponential growth chart.) In 1967, as part of one of the last big experiments8 he designed before his death, Barrow put his team of technicians to work on five hundred more books — this time imprints published between 1800 and 1899. Some of these books would be nice to have now — an 1817 edition of Marmontel’s Les Incas, ou la destruction de l’empire du Pérou (with plates); Bayard Taylor’s A Visit to China, India, and Japan in the Year 1853 (1855); Jones’s Medical Electricity (1895); the 1857 edition of William Cowper Prime’s Boat Life in Egypt and Nubia; Secret Journals of the Acts and Proceedings of Congress (1821 ed.); and Margaret Oliphant’s Makers of Venice. But Barrow, ever the dissector, had them snipped and clamped into the fold testers, which must have been waggling away into the wee hours, since the experimental regime demanded thirty strips per book — ten strips cut across the lines of print, ten cut from inkless paper, and ten cut parallel to the lines of print — not to mention another eight strips per book to be clamped and torn in the Elmendorf Tester. Barrow, who sometimes seems really to despise paper, found the “debasement of quality” to be “pervasive” in the books published between 1850 and 1869, but it “reached an all time low” at century’s end. (The descriptive writing here is probably Verner Clapp’s, not Barrow’s; Clapp’s literary assistance9 at times amounted to ghostwriting.) Over two thirds of the 1870–1899 group endured one fold or less, which meant they were, he said, “not suitable for regular library use.”
All these test books, along with the five hundred from 1900 to 1949, and assorted other lots (including seven books10 printed between 1534 and 1722, “all of which were in excellent condition after several centuries of use”), have now disappeared from view — temporarily, one hopes. The Barrow lab closed in the seventies; the books reportedly went to the Library of Congress. Bill Minter, the encapsulator, observed to me that even in their mutilated state these books would be interesting to study now: using Barrow’s baseline fold data, we might measure whether the paper had become appreciably weaker after thirty further years of natural aging, and how well deacidified paper held its deacidification, and we could get a better sense of what Barrow meant by “not suitable for regular library use.” Minter proposed to Chandru Shahani that the Library of Congress do some experiments on the Barrow test specimens (“For the past ten years I’ve been talking about this!” Minter says), but to judge by his research, Shahani is not terribly interested in the actual aging of paper — he remains fascinated by laboratory ovens, and by the possibility of developing an improved and simplified artificial-aging test.
Many of the books may be gone, but their quotably quantified test results live on in the hearts and bibliographies of preservation managers, who by the late seventies began to have fantasies of sampling the paper in their own collections, in order to see how catastrophically degraded and grant-gettably reformattable it was. Most libraries don’t have MIT Fold Testers, though — and anyway you wouldn’t want to be cutting strips out of your library’s books with Barrow’s abandon; you need something quicker and less extreme, albeit variable and imprecise. You also want something that undergraduates and other low-wagers can do with minimal training. And that’s how the library world settled on the double-fold test.
Anyone can do it. Open a book to a random page and fold its lower right corner in toward you, forming a triangle against the paper, until you feel it crease under your thumb. Then fold it back in the opposite direction until it folds against the far side of the page. That is one double fold. Do that until the paper breaks, or until you reach some stopping point, as specified by your library’s preservation department — one double fold, two, four, five. Double folding may seem oddly familiar to some, for it is how kindergarteners are taught to divide a piece of paper without scissors. Now, however, it is used to survey research collections in order to determine their “usability” and hence their fate.
“Usable,” as it happens, is another piece of specialized preservo-vocabulary. A unusable book is not a book that you can’t use. “An ‘unusable’ record,”11 wrote Gordon Williams in 1964, “is one already so deteriorated the paper breaks when folded once. A ‘usable’ record is one with obvious signs of past use but that might be expected to last at least another twenty-five years if untreated and stored under average library conditions.” But that is just one of dozens of definitions and regional variations. Indiana University defines a brittle book as one that doesn’t survive three double folds followed by a gentle tug, while an extremely brittle book is one whose page “breaks off12 in your hand when folded once”—in such cases, “it is often best to withdraw, replace or reformat the item.” At Northwestern, staff members are urged to do the “four corner test”13 for brittleness, “because we might end up reformatting the item,” but they are advised not to pull the corner. At Cornell and Berkeley, a brittle book is one that doesn’t withstand one double fold. At Johns Hopkins, when I called, it was one-and-a-half double folds; i.e., “three half folds.” The Library of Congress also uses three half folds. Ohio State defines brittleness as a paper’s breakage “when a lower corner14 is folded back and forth four (4) times (the ‘two-double-fold test’).” It’s two double folds at the University of Maryland, too, “at a width of no more than 1/2 inch,” followed by “a very gentle tug”;15 books that fail this test are “in jeopardy when anyone16 simply turns the leaves.” David Lowe, who manages an NEH-funded microfilming project at Columbia University,17 explained his library’s procedure to me: “Three-eighths of an inch from the corner you fold once, then back under for a single double fold, and then try to tug gently. Then single folds after that up to a total of four single folds — so only twice, after the double fold. Four is the max. If it withstands four, we don’t torture it any more.”