R - S

- Rising Spaces and Quads [see Tanselle, ‘Treatment’ cit., pp. 41-43: “Impressions from material not meant to print”]. The sight of a space between two words rising and catching ink is familiar to most bibliographers, and is worth noting in collation, since it can often distinguish a different state of the sheet. A somewhat unusual case in which type placed inside woodcut tends to rise when the wood swells is described in Neil Harris, ‘Rising quadrats in the woodcuts of the Aldine Hypnerotomachia Poliphili (1499)’, Gutenberg Jahrbuch (2002), pp. 158-167.

- Set-Offs (or Off-Sets) . It occasionally happens that in an early printed books, we find a set-off or the shadowy, back-to-front image left by the contact with another freshly-inked sheet. In the majority of cases it comes from the same book, though rarely from the same sheet, but at times from another book altogether. The potential of set-offs as a mechanism for dating and identifying the printer of books produced at the same time in the same shop, especially in Florentine Fifteenth-Century output, was first pointed out by Roberto Ridolfi, ‘Incunaboli contrastampati. Nuovi sussidi per l’attribuzione e la datazione dei paleotipi’, La Bibliofilìa, vol. 51 (1949), pp. 131-144, and later studies have followed along the same lines; see Adolfo Tura, ‘Sull’anno di stampa di due edizioni di Ripoli’, La Bibliofilìa, vol. 100 (1998), pp. 43-46; Edoardo Barbieri, ‘Tra filologia dei testi a stampa e storia del libro. Ridolfi, Cicerchia e le “controstampe”’, in Bibliografia testuale o filologia dei testi a stampa? Convegno in onore di Conor Fahy, Udine 24-25-26 febbraio 1997, a cura di Neil Harris (Udine, Forum, 1999, pp. 35-58). The unspoken assumption in all these studies is that the set-offs are produced by an accidental contact between sheets, but two examples of Fifteenth-Century books are known which present a very high concentrations of set-offs in the same copy. The first is an exemplar of the Vita S. Augustini and of the Defensorium ordinis Heremitarum S. Augustini by Ambrogio Massario de Cora (Rome, Georgius Herolt, 1481 in the first instance; unsigned in the second but from the same press) in the Gutenberg Museum at Mainz; see Adolf Tronnier, ‘Zwei kleine Beiträge’, Gutenberg Jahrbuch (1950), pp. 121-127. The second is the only large-paper copy so far known of the Aldine Hypnerotomachia Poliphili (1499) in the Archbishop’s Library at Udine, in which 330 set-offs appear (285 of them identified), many of them in layers of up to seven in a single side of the same sheet; see Neil Harris, ‘L’Hypnerotomachia Poliphili e le contrastampe’, La Bibliofilìa, vol. 100 (1998), pp. 201-251, also issued in the centennial volume Anatomie bibliologiche: saggi di storia del libro per il centenario de «La Bibliofilìa», a cura di Luigi Balsamo e Pierangelo Bellettini (Firenze, Olschki, 1999). These two items show inequivocably that the set-offs are a deliberate side-effect of some process in Fifteenth-Century printing about which we have no knowledge. With reference to the manualistic literature of the Nineteenth century, Harris suggests that these were originally protective smut-sheets, used to protect vellum leaves on which the ink was slow-drying; the press-workers took care however to print these sheets, which would otherwise have been discarded, so as to obtain an extra copy, perhaps as a prerequisite for their labours. Though the Udine copy of the Polifilo was rebound in the Eighteenth century, it conserves an original end-paper set-off by other pages in the work, showing that it was originally made up and bound with waste materials from the Aldine shop.

- Setting by Formes. Though the validity of the model proposed by Hinman and others for the organisation of the printing shop has been cast into doubt by McKenzie, the need of early printers to coordinate efficiently the work of the compositors and the press-men remains unquestioned. In particular the former had to garantee a steady flow of formes ready to be printed by the latter. Students are invited to work out the following problem, if necessary by folding a few sheets of paper. If a printing shop has the equivalent of eight pages of type in the case, begins to set text for a folio format gathered in eights, and proceeds seriatim (i.e. as one reads), which will be the first forme to be completed and ready? Likewise, if it sets out to print a quarto format gathered in eights (i.e. two leaves, one folded inside the other), has the equivalent of ten pages of type in the case and again proceeds seriatim, which forme will be ready first? Once you have correctly worked out the answer, it should be clear that there must be a better way of doing it. The alternative solution is in fact to set by formes, in other words in the order in which the pages are printed and not in that in which they are read. The manuals of Moxon, Fertel and other early printers contain precise instructions about how to cast off copy, both to determine the length of the book in the transition from manuscript to type and, though this is not stated explicitly, to allow the compositors to set by formes (this application is described, however, in the Spanish manual by Paredes). The best known case of setting by formes is described by Hinman, The Printing and Proof-Reading of the First Folio of Shakespeare cit., vol. 1, pp. 47-51, who shows that in the 1623 First Folio work began with the innermost leaf of the six-leaf gathering, i.e. in the order 3v.4r, 3r.4v, 2v.5r, 2r.5v, 1v.6r, and 1r.6v, generally with one compositor doing the backward sequence of 3v to 1r and another doing the onward sequence of 4r to 6v. Expert bibliographers can recognise signs of difficulties caused by setting by formes, though the only sure proof is that provided by following a pattern of damaged characters, which sometimes show that it was physically impossible for the book to have been printed seriatim. Though it would be useful to have a critical listing of work demonstrating setting by formes (or the contrary), some further indications are provided in David J. Shaw, ‘Setting by Formes in Some Early Parisian Greek Books’, in Book Production and Letters in the Western European Renaissance: Essays in Honour of Conor Fahy, edited by A. L. Lepschy, J. Took, D. E. Rhodes (London, The Modern Humanities Research Association, 1986, pp. 284-290); Frans A. Janssen, ‘Some Notes on Setting by Formes’, Quarendo, vol. 16 (1986), pp. 191-197.

- Skeleton Formes. The term sounds gruesome, but it basically means the materials placed around the letterpress in the forme, including the running-titles. The Oxford English Dictionary credits the first use of the term to Bowers: “when the term skeleton is used it will indicate the imposed cross-bars, furniture, and running-titles of a forme”; see ‘Notes on Running-Titles as Bibliographical Evidence’, The Library, s. IV, vol. 19, 1938, pp. 315-338.

- Slip (Pastedown, Papillon). At times, rather than go to all the trouble of setting type and printing new sheets of paper for a trifling correction, printers used to stamp the correction on a piece of paper, cut it out and glue it onto the sheets. Such corrections are bibliographically significant, since at the least they constitute a variant of state and at the most, when the change is made to the imprint, a variant of issue.

- Standing Type [see Tanselle, ‘Treatment’ cit., pp. 43-51: “Variation in furniture width”]. The very small quantity of type available to early printers meant that they were usually loath to keep a forme standing any longer than was strictly necessary, but there are cases, especially title-pages, in which the same basic setting can be shown to be used several times; see Fredson Bowers, ‘Notes on Standing Type in Elizabethan Printing’, Papers of the Bibliographical Society of America, vol. 40 (1946), pp. 205-224; Robert K. Turner, ‘Standing Type in Tomkis’s Albumazar’, The Library, s. 5, vol. 13 (1958), pp. 175-185; William B. Todd, ‘Recurrent Printing’, Studies in Bibliography, vol. 12 (1959), pp. 189-198.

- Statistics. No truly reliable statistics have yet been produced regarding early book production, but see Paul Needham, ‘Counting Incunables: The IISTC CD-ROM’, Huntington Library Quarterly, vol. 61 (2000), pp. 456-529. Over the years a large number of estimates have been made relating to Fifteenth-Century output, which have generally oscillated between thirty and forty thousand editions (at the time of writing ISTC stands a little over 26,000 entries, which therefore represent certainties), but the authors have usually failed to make it clear whether they refer to all production, including editions that have not reached the present day, or just the survivors. Since it seems reasonable to suppose that something between 20% and 40% of the editions printed in the Fifteenth century, excluding ephemera, have disappeared in every single copy, the distinction is an important one. Any figures given for the Sixteenth century onwards are largely guesswork, since the only European country for which reliable figures about surviving production are available is Britain, a marginal producer (see § 6 above). Attempts have been made for Italy to provide estimates based on the holdings described in the Short-Title Catalogue of Italian Books (1958) of the British Library, but the calculation takes no account of the way the collections were formed under the influence of bibliophily of the late Eighteenth and early Nineteenth centuries, so that the final outcome is dangerously misleading, see Amedeo Quondam, ‘La letteratura in tipografia’, in Letteratura italiana: II. Produzione e consumo, a cura di Alberto Asor Rosa (Torino, Einaudi, 1983, pp. 555-686), and the comments on the figures produced in Neil Harris, ‘Appunti per una logica del catalogo delle cinquecentine’, in Le cinquecentine della Biblioteca Panizzi. Catalogo, a cura di Eletta Zanzanelli, Valter Pratissoli (Reggio Emilia, Biblioteca Panizzi, 1995), pp. xi-xxiv. It is generally safe to assume that statistics about early printing are usually produced by people who know nothing about statistics or nothing about bibliography or, quite often, nothing about either.

- Survival of Books. It has long been an ambition of mine to write a bibliography of all the editions of books produced by printers of which the entirety of the copies has been lost. It bodes to be a sizeable work, since, including an adequate historical introduction and a few paragraphs for those editions about which we have information from documentary sources, several hundred blank pages will be required to furnish a complete listing. One book along these lines, albeit without blank pages, is Donald Wing, A Gallery of Ghosts. Books published between 1641–1700 not found in the “Short-Title Catalogue” (New York: the Index Committee of the Modern Language Association, 1967), which lists some 5,000 titles found in secondary sources, including booksellers catalogues, and which were not discovered in the two hundred libraries whose holdings were trawled for his catalogue.

In very recent times bibliographers and historians of the book have begun to realise that survival really is a major issue or, to borrow a concept from Terry Belanger, we can write history from the point of view of The Titanic or from that of the iceberg; see Thomas R. Adams-Nicolas Barker, ‘A New Model for the Study of the Book’, in A Potencie of Life: Books in Society. The Clark Lectures 1986-1987, edited by Nicolas Barker (London, The British Library, 1993, pp. 5-43); David McKitterick, ‘The Survival of Books’, The Book Collector, vol. 43 (1994), pp. 9-26. Any and every scholar of the book has to recognise that the mysterious, unknown force that has totally destroyed millions upon millions of tomes lies within our own selves, in the act of reading or, more simply, of use. The important question therefore is not to understand why so many have disappeared, but why the few that have survived actually managed to do so. Obviously there are certain categories of books – school books, books of worship, and so on – that are heading for destruction from the moment of purchase, but there are also perhaps general factors that are valid for all printed output. The first and most obvious element is sheer size, expressed not only by the size of the original sheet and the format, but also by the number of leaves in the book which determine the thickness. The second major force is language: when the same book is published both in Latin and in a vernacular tongue, copies of the latter are significantly more difficult to find in today’s libraries; see Jean-François Gilmont, ‘Livre, bibliographie et statistiques: à propos d’une etude rècente’, Revue d’histoire ecclésiastique, vol. 65 (1970), pp. 797-816; Idem, ‘La diffusion et conservation des editions de C. Scribani’, Revue d’histoire de la spiritualité, vol. 53 (1977), pp. 267-274. The third significant factor is time or, more properly, the time required by the new revolutionary way of making books to take effect on the reading and book-owning public. Although by the end of the first quarter of the Sixteenth century books had become considerably smaller and less robust, the average survival rate unquestionably improves. Several reasons can be advanced: the much larger quantity of books in circulation meant less wear and tear on single items; better forms of storage, in particular a new way of placing volumes upright on bookshelves, which seems to have appeared some time around 1530; and, most important of all, a new reading public better versed in ways of looking after their books.

In the evaluation of survival statistics the scale of press-runs also has to be taken into account, but not, as one might think, because a larger run guarantees a greater survival rate. In an important article, unfortunately little known, perhaps because published during the Second World War, O.M. Willard compares known press-runs to survival rates in the STC and concludes that, as a general rule, survival is in an inverse ratio to the number of copies printed, i.e. if a small number of copies are printed, a high percentage of them will survive; if a large quantity is produced, few of them will be findable in today’s libraries; see ‘The Survival of English Books Printed Before 1640: A Theory and Some Illustrations’, The Library, s. 4, vol. 23 (1943), pp. 171-190. At first the whole proposition seems a paradox, but mature reflection shows its wisdom and, in my opinion, it deserves to be enshrined as a fundamental bibliographical principle (albeit with some exceptions, for instance, by Renaissance Venice standards Aldines had high press-runs and also survive in large numbers). Books with small print-runs tend to be more high-brow, better executed on superior paper, and above all cost more, since the publishing house includes in the price the longer warehouse time required to dispose of the stock of copies; those with large print-runs are aimed at a wider readership, not made up exclusively of scholars, are less robust in their manufacture and are less looked after, because they are known to have cost little in the first place.

Nevertheless bibliography has to come to terms with the fact that a huge number not only of exemplars but also of editions have been lost and have in some way to be accounted for. In 1932 Ernst Consentius suggested that the total number of ‘lost’ imprints belonging to the Fifteenth century could be calculated by compiling a graph of all the editions surviving in one copy, two copies, three copies, four copies, and so on. The projection of the graph should provide an approximate indication of the editions not surviving in any copies at all; see ‘Die Typen und der Gesamtkatalog der Wiegendrucke: eine Kritik’, Gutenberg Jahrbuch (1932), pp. 55-109. This same principle of zero-graphing is examined and applied in a misleadingly titled, since it fails to signal its relevance to book statistics, article by Neil Harris, ‘Marin Sanudo, forerunner of Melzi’, La Bibliofilìa, vol. 95 (1993), pp. 1-37, 101-145, vol. 96 (1994), pp. 15-42, which looks at an early list, probably written in 1528, of Italian chivalric editions of which half are unknown today and the others mainly survive in single or double copies. Whatever reservations can rightly be expressed about zero-graphing, in particular about the reliability of copy-counts, even in state-of-the-art repertories such as ISTC and ESTC, an indication based on a statistical projection is always going to be more accurate than sheer guesswork.