Franca Petrucci Nardelli
Translated by Marinella Mezzanotte

Originally published as: Franca Petrucci Nardelli, “Funzione della legatura e modi di conservazione libraria,” in Franca Petrucci Nardelli: Fra stampa e legature, ed. Chiara Carlucci (Manziana: Vecchiarelli Editore, 2000), 219-26, after a paper presented at the congress Il luogo dei libri. Lo studio-la biblioteca held in Udine, 12-13 November 1993 (part 1).

Between the latter half of the fourteenth century and the first half of the sixteenth, some structural changes occurred in bookbinding, taking the practice from the medieval to the modern. In earlier periods, change mostly took place in how sections were sewn together or bands attached to backing boards. Later, even though sewing methods changed – still speaking from a structural viewpoint – they did so on grounds of aesthetics rather than functionality. The aforementioned time period ushered in a multitude of changes. Along with new ways of attaching bands to boards, overhangs made their first appearance. Studding decreased significantly, metal clasps were replaced by ribbons, and cardboard, or rather pasteboard, began to be used instead of wood as a backing material. Chains remained in use, but only for certain books and in certain libraries. Decoration was directly influenced by Islamic and mudéjar style, later adopting autonomous models that would develop alongside other arts of the time in line with Renaissance tastes.

Further changes in how bands were attached to wooden backing boards can first be observed in the early fifteenth century. Until then, bands had always been inserted through grooves drilled directly through the edge of the board from the outside in. The band ends would then be fixed to the inner board. From approximately 1400, in a process that would come to a conclusion in the latter half of the century, bands began to hug bookboards from the outside, before going through them to be fixed to the interior side, with or without further passages. This change, important for dating bound volumes, concluded the search for a way of solidly attaching the pages to the boards – a traditional weak spot of bound volumes. As such, it seems unconnected to how books were read or shelved.

Somewhere around the 1480s, backing boards began to be made of pressed paper pasteboard rather than wood. This innovation appears to be linked to the increase in the number of books requiring binding, caused in turn by the establishment of printing technology. Even once books were “artificialiter scripti” [written by means of artifice] in large numbers, they still had to be bound by hand, like handwritten manuscripts. Books also became smaller in format, which meant cover materials could be attached to a less solid, less heavy backing.

This certainly had an impact on both reading and shelving. Obviously, smaller, lighter books are more easily handled, becoming “hand books” as opposed to “desk books”.[1] As for shelving, the adoption of pasteboard backing can be linked to the decreasing use of studding. Studs did not fall into disuse due to the adoption of pasteboard (though it would have been trickier to apply them);[2] rather, their disappearance is linked with changes to the shelving position of books. The function of studs was to protect books resting horizontally on backing boards, preventing abrasion of the cover material from contact with the shelf. Studs did, however, risk damaging books stored alongside them, vertically or horizontally. These two phenomena do point to a change in shelving position, but it would be highly simplistic to state that as a rule, bound volumes with pasteboard backing and no studs, dating from 1480 onwards, equaled vertical shelving.

We should instead bear in mind that bound volumes with pasteboard backing could not have withstood the application of a chain; similarly, they gradually rejected metal clasps, eventually abandoning them altogether in favor of ribbons. This introduces a further aspect of binding with an impact on shelving. As we consider the presence or absence of chains, it becomes evident that any attempt to determine a single specific point when books stopped being shelved horizontally and were instead shelved in the now familiar vertical position does not reflect the diversity of historical situations. Not only do we observe different dates in different geographical regions, but also conflicting dates in the same regions, depending on the storage location and the function of books and libraries. Libraries at universities and large monasteries belonging to mendicant orders – collections created and developed for long study – were known to divide books into two categories: the less valuable, usually replaceable and loaned out to students and professors, and works of historical and commercial value to the institution. The latter went from being stored in locked chests or armarii to being placed together in one room and chained. Initially, these books were laid on a board at an angle, chained to stands or lecterns, with shelves eventually appearing above and/or below; later, they were shelved vertically above the lecterns, fore-edge facing outward. They were eventually shelved in bookcases, some of them still chained. This process, greatly simplified for the purposes of this presentation, has been documented and studied mostly, though not exclusively, for the libraries of Oxford.[3]

The site of a chain fastening on a bound volume would depend on how it was to be shelved. On books to be placed on an inclined stand, the fastening would be fixed to one of the boards, usually the back face, on either the tail or head square. The other end of the chain was attached to a horizontal bar at the tail or head of the volume depending on if the volume was to be placed below or above the stand. On books shelved vertically with the fore-edge facing outward, the chain was fastened to the front square of the front face. Traces of the chain’s fitting can therefore give clues about a book’s historic shelving position.

The first library designed and built for vertical shelving in wall-adjacent bookcases is at El Escorial, completed in 1584 and immediately followed by others, including the Bodleian in Oxford, which was expanded in 1610-1612 with the addition of two large wings purpose-built for this shelving system.

Despite the presence of chained books in the two new wings, the system itself came to exclude chaining, which usually required a seat and a lectern close to the chained volume. By excluding in situ consultation and reading, vertical shelving in wall bookcases also excluded the use of chains. In turn, the elimination of chains from vertically shelved books put an end to fore-edge-out shelving, a practice that arose directly from chain fastenings. Twenty-five years after El Escorial was built, the books in the Ambrosiana, Milan, were housed in wall bookcases, shelved vertically according to format, without chains and with spines facing outward. For that matter, when Gabriel Naudé was advising readers how to set up their own library in Advis pour dresser une bibliothèque (1627), he and his readers seemed perfectly comfortable placing books in a vertical position, spines facing outward, on shelves that hid the entire wall, hinting at the recently outmoded fashion of placing books on desks.[4] Modern shelving had already come into its own. There is, however, no documented evidence of vertical shelving in bookcases not adjacent to the walls being practiced in Italy or France as it was in England.

One of the oldest and grandest stately libraries of Italy was, perhaps, the long-dispersed one belonging first to the Visconti family, and later to the Sforza. The reconstruction of this library has been undertaken, as far as possible, by Elisabeth Pellegrin.[5] As far as I know however, no one has spoken about the physical housing of the books on the original site, located in the Castello Visconteo, Pavia.

Here, I will anticipate the conclusions of a short paper I have written on this topic. A study of the library’s book display has only been possible for the Sforza era, even though we are well acquainted with its Visconti-era status. As a typical state institution, its display reflects the customs of the time, acting as a political and diplomatic tool, rather than being used for reading, let alone writing. The books, therefore, were above all “well displayed”. The library comprised four spaces: a small entrance hall led to the introitus, which contained almost half the total number of codices. The corresponding exitus held fewer than one hundred codices. In a fortified tower between the introitus and exitus was the main room visitors would be shown into. It measured over one hundred square meters, had two large windows, and was decorated and hung with paintings. The room housed a series of exhibits such as globes, clocks, and even relics, along with books and archival documents. Some of the books were displayed on plutei[6] and others on the four shelves of open cupboards, where they probably rested at an angle. First-person accounts indicate at least some of the books were chained; nothing can be inferred from the bound volumes themselves, since they have all been lost.[7]

We can be certain that the books in the Malatesta Library, Cesena, were chained, because they still are. Built in 1452, this library’s shelving is an instance of a “lectern-system” in Clark’s typology.[8] The Malatesta is a prime example of a library designed and built for promotional ends by a Renaissance nobleman, differing in its purpose from libraries dedicated to religion and scholarship. Distant as it may have been from the more lively (and transient) study spaces of university libraries, and though not itself part of an academic institution, it bears strong analogies with the most valuable section of any such library, the books chained and still.

In the Malatesta library, the books are laid horizontally on the shelf below the bookstand, or at an angle on the stand itself, resting in both cases on the front face; the back face of many of them is labelled with the title, and bears the chain fastening on the tail square. The same display can be found in the Medicea Laurenziana, Florence, created by Michelangelo and completed in 1571; but it seems pointless to say much more about it, since it is in my opinion the great artistic fulfilment of the idea of a library, rather than an actual library built to be used as one.

I think we can therefore accept the following as an established relationship: in Italy, the presence (or trace) of a chain on the head or tail square of the back face, and the presence of studs and wooden backing, point to horizontal or semi-horizontal shelving. Conversely, pasteboard backing and an absence of studs and chains may not ipso facto confirm the vertical shelving of a book.

So far, I have not touched on one structural innovation in binding mentioned at the beginning of this presentation, nor on private shelving practices, both of which pose rather thorny problems.

The first appearance of the overhang in bound volumes dates to the late fourteenth century. Common sense tells us that its absence could do no damage to a horizontally shelved book, but when a book was shelved on an inclined stand, its weight would rest partially on the tail of the boards and text block. A vertical shelving position would cause the full weight of the bound volume to rest on the tail edges of the case and text block. The damage once avoided by fixing studs to a book’s boards to isolate them from the shelving surface would now be inflicted on the tail of a book, even affecting the margins of a volume that had not been bound tightly enough. The overhang lifted the body of a vertically bound book, isolating it from the shelving surface – a function previously served by studs.

In short, the direct relationship between the overhang and vertical or inclined shelving has come to light through practical observations rather than documentary or explicit evidence from a specific point in place and time. An assertive statement could only be made after further systematic observation, and perhaps research into any indirect evidence of this relationship. In any case, the Malatesta codices, which were created specifically for that type of library and that type of shelving, feature studs, metal clasps, and overhangs.

As for shelving, so far I have only spoken about libraries for public use. How can we connect the technical innovations in binding that we have examined so far to privately owned books, often few in number?

From the little we know, privately owned books – particularly small collections that did not amount to a library proper – would be kept in closed shelving units. In private homes, just as in public institutions prior to the adoption of bookstands, books would be shelved in cupboards, or rather chests, along with other possessions. This shelving practice, at risk from oil lamps and inquisitive children, continued even after institutional libraries had adopted other ways of displaying their books.

Here, I am not thinking of the numerous books owned by private scholars, but rather of the two, five or ten books perhaps owned by a minor merchant or artisan. Such books certainly never had chains, and rarely studs. They would be very frequently stored with “family books”, many of which would have come with a type of binding that I tend to call archival, although it should probably be defined as mercantile or documentary. These bindings were often soft, with light or non-existent backing, at other times hard, and differ in decoration and structure from the more familiar library bindings.

Let us examine the binding of a manuscript housed in Chicago’s Newberry Library, dated prior to 1360.[9] It is a ledger and memorial book kept by a member of the Albizzi family, with a portfolio binding (i.e. with a flap) consisting of two layers of leather. The outer layer is dyed pink, the inner white; they are sewn rather than glued together. Contrary to what we often find in this type of codex, the body of the book is sewn onto the (cut) bands, whose ends are slotted into incisions in the layer of leather lining the front cover; likewise, the four ends of the head and tail bands slot into the lining by way of small incisions, thus remaining concealed between the two layers of leather. The same technique was not used at the back, where the ends of the bands were left loose between the inside of the back cover and the book itself.

The ends of the bands slotted into the front cover are stitched in place with thin, twisted strips of illuminated parchment. The stitches appear on the inner cover, crossing over in a pattern at once decorative and functional, joining the bands (sewn onto the body of the book) and the cover, while also serving to attach reinforcing strips to the outer cover. The central strip, longer than the others, has a plain metal buckle. Two stitches crossing over on the exterior of the spine, corresponding with the bands, helped to keep the book and cover together: only a few traces of these remain. The first records in the book date to 1339, the last to 1360. It was bound after this date. While in use – unusually, for this and similar family books, “use” refers to writing rather than reading – the sections were probably not sewn together; they would have been kept in a sheet of thick, stiff parchment, its bottom leaf folded over the top to resemble a wallet. This feature was retained when the time came to bind the book, remaining between the body of the book and its cover.

On the front cover, below the flap (and therefore invisible when the book is closed), the motto “PAX” [peace] is written in ink. On the back cover is the owner’s name, Pepo, parallel to the right fore-edge, framed by a rectangle. I can only deduce that, with no hard backing, this codex must have been shelved horizontally, back face up, with similar books stacked underneath or on top of it. It would have stood out thanks to the reddish pink leather binding among books of other colors (cross-referenced through internal records), while the positioning of the owner’s name meant that it could be identified without disturbing the books piled on top of it. We see archival (or mercantile) binding on comparable volumes but with a different structure: they may have been hard bound, or bound prior to use.[10]

Family books certainly belong to a category of their own, but they are books nonetheless, and the same questions arise about the way they were shelved as for other books. How and where did people keep these “books of reasonings and memories”?

The data gathered by A.F. Verde’s study[11] of the official register of wards of the city (1471-1503) indicates that such books would be kept in escritoires[12] and armarii [cabinets] – one “picoletto” [rather small] and another with two “serrami” [locks][13] – as well as strongboxes[14] and chests.[15] Where the type of binding is specified, boards are never mentioned; if anything, the books are described as “chovertati con charta pecora” [covered with vellum]. [16] They would be shelved with little ceremony, as real books were – and by “real”, I mean books that were neither ledgers nor family memorial books, but “for reading”, whether for work or study,[17] and were usually bound with wooden backing boards. Not even “women’s little books” commanded more attentive shelving, despite being very attractively bound. Other than three or four with leather bindings, most of these very pretty (not to say frilly) books were covered with sought-after materials such as green, crimson or gilt brocade,[18] Alexandrine,[19] black, red, green[20] and crimson[21] satin, white or green brocatelle,[22] turquoise brocade,[23] azure damask,[24] damask with bouclé metal threads,[25] and purple velvet.[26] Most had silver clasps, but only a few were decorated with pearls; one was embroidered in gold thread.[27]

We know these to have been not so much books as adornments for wealthy Florentine women. It is therefore surprising that the wife of the famous Prato merchant Francesco Datini, Margherita (who learned to write as an adult so she could correspond with her husband), was criticized by Lapo Mazzei for the unworthy “copritura” [cover] of her “book of Our Lady”: in the notary’s words, “shame on those who keep in their homes books with deplorable covers, worldly though they may be”.[28]

Whether treated as books or objects, I like to think that women’s books would have been kept with some care, together with other books (whether written in the home or elsewhere) and family possessions, perhaps wrapped in one of the sashes that we find listed among the contents of bridal chests.

There were therefore commonalities between the shelving and storage of even very modest family book collections, and those belonging to great educational or aristocratic institutions, before or during the adoption of the lectern system. Family shelving practices remained in use for far longer than the lectern system, always retaining their miscellaneous nature. The few books a family might own lacked not only a dedicated room in the home, but also their own specific shelf. Between the great institutional libraries and those consisting of minor or small family collections, there is a third type of library, belonging to private scholars, which differs from both in size and function and must therefore have adopted shelving practices of its own. One possible way of researching such libraries is the analysis of iconographic evidence. While this is one of the most enjoyable forms of investigation, it is also one of the most treacherous, due to the uncertain documentary value to be attributed to the various sources of evidence.[29]


[1] Armando Petrucci, “Alle origini del libro moderno. Libri da banco, libri da bisaccia, libretti da mano,” Italia medioevale e umanistica 12 (1969), 295-313, translated in this issue as “The Origins of the Modern Book: Desk Books, Pouch Books, and Paperbacks.”

[2] Studs with a primarily aesthetic function can be found on books well into the seventeenth century.

[3] Burnett Hillman Streeter, The Chained Library. A Survey of Four Centuries in the Evolution of the English Library (London: MacMillan, 1931).

[4] Gabriel Naudé, Advis pour dresser une bibliothèque (Paris: François Targa, 1627), 149.

[5] See, for instance, Elisabeth Pellegrin, La Bibliothèque des Visconti et des Sforza ducs de Milan au XVe siècle (Paris: Service des Publications du C.N.R.S., 1955) (editors’ note).

[6] A large stone or wooden balustrade dividing a room (editors’ note).

[7] Franca Petrucci Nardelli, “La biblioteca Sforzesca. Ubicazione e disposizione del materiale librario,” La Bibliofilía 97 (1995), 21-33.

[8] John Willis Clark, The Care of Books. An Essay on the Development of Libraries and their Fittings, from the Earliest Times to the End of Eighteenth Century (Cambridge: Cambridge University Press, 1909).

[9] Add. 26 (27). Cf. Paul Saenger, A Catalogue of the pre-1500 Western Manuscript Books at the Newberry Library (Chicago: The University of Chicago Press, 1989), 49; see plates 18 and 19.

[10] Franca Allegrezza, “La diffusione di un nuovo prodotto di bottega. Ipotesi sulla confezione dei libri di famiglia a Firenze nel Quattrocento,” Scrittura e Civiltà 15 (1991), 252, 255.

[11] Armando F. Verde, “Libri tra le pareti domestiche,” Memorie domenicane 18 (1987), 1-225.

[12] Verde, “Libri tra le pareti domestiche,” 177, 115.

[13] Verde, 66, 173.

[14] Verde, 199.

[15] Verde, 201.

[16] Verde, 82, 177.

[17] Verde, 97, 136, 137, 176.

[18] Verde, 46, 96, 139.

[19] Verde, 79. The exact nature of alexandrine is unclear: it may refer to wool or silk woven in Alexandria, Egypt, or fine linen or cotton from France and Switzerland treated to resemble silk (editors’ note).

[20] Verde, 59, 176, 71.

[21] Verde, 110, 139.

[22] Verde, 171, 60.

[23] Verde, 108.

[24] Verde, 61.

[25] Verde, 115.

[26] Verde, 93.

[27] Verde, 139.

[28] Lapo Mazzei, Lettere di un notaro ad un mercante del secolo XIV, ed. Cesare Guasti (Florence: Succ. Le Monnier, 1880), II: 181.

[29] I put myself forward to illustrate a few cases relevant to this topic on a future occasion. [The forthcoming work mentioned here is also included in this issue of Lingua Franca as “Image of the Book” (editors’ note).]