Mario Carpo

 Publication History:

“How Do You Imitate a Building That You Have Never Seen?  Printed images, ancient models, and handmade drawings in Renaissance architectural theory.” Zeitschrift für Kunstgeschichte 64, 2 (2001): 223-234

The text posted here is an earlier draft and may be different from the published version. Please only cite from copy in print   

In his Life of Brunelleschi, Vasari recounts an anecdote that must have struck the imagination of the Florentines at the beginning of the fifteenth century—to the point that the memory of it was still alive in town when Vasari recorded it in writing, more than one century later.  Somewhere in Florence, Donatello had praised the beauty of a sarcophagus that he had seen in Cortona.  Moved by his words, a Florentine artist went to Cortona to see it with his own eyes; he made a drawing, and when he came back, he showed it around.1  This story may seem uneventful today—much as it must have seemed at the time of Vasari.  However, it would have been exceptional at the time when it supposedly took place.  If Vasari’s story is true, and one might have some doubts, here we have a situation in which—in Florence, at the time of Brunelleschi and Donatello, and at the dawn of the humanistic rediscovery of classical antiquity—visual information on a work of art was being recorded and communicated to the general public in a visual format.2 

Elated by Donatello’s recital, his follower went out of his way to draw a picture of the far-away sarcophagus.  Back in Florence, instead of climbing on a stool in the middle of a busy marketplace to tout the glories of what he had seen, he showed his drawing.  This speechless art of describing, unusual as it must have seemed in the famously vocal Florentine environment, was the harbinger of a major cultural change.  For centuries in the West, images of works of art and of mirabilia of different kinds, seen or imagined, had been transmitted primarily by word of mouth, orality and memory having been occasionally superseded or complemented by alphabetical writing.  Drawings did not participate in this process, and ample evidence proves that when they did, it was by accident: as everyone knew at the time, drawings could not, and should not, be relied upon.  In a word, they did not matter.  Visual forms were to be described by words, not by pictures.3

As daily experience can prove, the translation of images into words may at times turn out to be a daunting or even an impossible task.  Many technical and scientific writers of antiquity, the Middle-Ages, and the early Renaissance, were confronted with a similar impediment, which in many cases effectively hampered or barred the way to the communication of essential visual information.  The problem of the transmission of artistic models in the Middle Ages was formulated almost a century ago by Julius von Schlosser, but no convincing answer to Schlosser’s questions has been suggested to this day.4  The peculiarities of the medieval notion of imitation were outlined by Richard Krautheimer in a pioneering essay of 1942:5 as Krautheimer remarked elsewhere, the medieval vision of antiquity was “almost emphatically nonvisual”.6  Indeed, medieval models, whether ancient or modern, were neither visual nor visualised.  Instead, they were verbal, and oralized.  Medieval artisans, builders, and even painters, were almost always working from models that they had never seen.  They imitated models that they knew only through hearsay.

The ancient and medieval aversion to the transmission of drawings was more than justified by some inevitable material conditions.  Likewise, ekphrasis—traditionally one of the hardest exercises of classical rhetoric—was as much a display of literary virtuosity as the logical cultural adjustment to a specific technical environment.  Before the invention of printed images, drawings were not easily reproducible.  Hence they were not used when and where easy reproducibility was required.  The written word was much more easily transmissible than pictures: a safer, cheaper, more reliable medium to convey, transmit, disseminate and, so to speak, broad-cast information in space and time.  Hence, when transmission and transmissibility were relevant, pictures were often translated into words.  For centuries, the alphabet was the standard code for information interchange. 

The advantages of the alphabet over other competing media were staggering, and unquestionable.  The alphabet is a machine for the digital recording of spoken words.  The same machine, if available, enables readers distant in space and time from the original writer to recreate the same sounds, or almost the same, which in most cases may also bring back to life the original meaning that was meant by the original words.  This machine is based on a code of some dozens of standardized graphic signs.  Literacy is relatively easy to master, and, to this day, it still performs well.  In a modern classroom, as well as in a medieval scriptorium, a reader dictates a passage, twenty scribes—or pupils—write it down, and at the end of the day we have a multiplication of almost identical texts.  They look different, but when replayed, that is, read aloud, they still contain a fair share of the content of the original text.  Alphabetical text can be dictated, written down, and transmitted verbatim.  Pictures cannot be transmitted verbatim, for the good reason that they cannot be broken down into verba, let alone into letters.  Copying a drawing is a risky venture.  It demands training and talent, and as a rule, the similarity between the archetype and its copy is never to be taken for granted.  It depends on the talent of the copyist, his or her motivations and mood, the weather, the time at his or her disposal, and last but not least on the complexity of the drawing. 

Pictures cannot be dictated, and in the graphic domain, fidelity to the archetype is an elusive notion.  There is no such thing as a spelling mistake in drawing.  Today we would say that the transmission of alphabetical text is digital, the manual copy of images is analogical.  As Walter Ong wrote some years ago, the alphabet was the first technologizing of the word.7  Until the invention of xylography, no comparable technology was available for the transmission of images.8  Images could be drawn; but drawings could not be transmitted in any controllable and verifiable way.  Hence, in many cases, drawings were avoided right from the start.  Vitruvius thought better of burdening his text with a complex iconography, because he knew that after the first copy such an iconography would have been abandoned—manipulated, or hopelessly deformed.  Renaissance authors, as well as more recent ones, have looked high and low for the lost illustrations of Vitruvius, and have invented the most bizarre theories to account for their disappearance.  But the illustrations of Vitruvius never existed, with the exception of nine or eleven elementary geometrical diagrams, mentioned in the text and first itemized by Philandrier in 1544.9  Pliny and Galen had already warned the authors of scientific texts not to use any sort of illustrations.  Pliny is categorical: the destiny of an image in a manuscript in unpredictable.  No one can tell how the next copyist may distort it.10  Faced with even more sensitive visual data, classical geographers resorted to some original scanning methods to translate their pictures into sequences of letters and numbers: in Ptolemy’s own words, this was necessary to reduce or circumvent the distortions and errors generated by the manual drawing and copying of maps.11  Many centuries later, Leon Battista Alberti, while advocating the use of drawings for architectural designs, insists and reiterates that his architectural treatise is not, and should never be, illustrated.  Copyists can already make a mess when copying simple letters and Roman numerals, Alberti complains; quite understandably, he decided to avoid illustrations altogether.  Alberti himself had, if we believe his word, profusely drawn buildings old and new, and he certainly had some interest for the medium of drawing, and some newfangled technologies related to it.  But when he conceived of a treatise that was to be transmitted to posterity, and possibly to eternity, he deliberately chose to rely on that good, old, reliable medium of all (historical) times—the alphabet.  Alberti’s architectural theories were to travel in space and time encapsulated in a digital file: verbis solis.12 

However, this anti-visual scenario had already begun to change, and change certainly was already in the air, at the time of Alberti’s writing.  Scholars have recorded a surge in the number of drawn model books, or pattern books for painters and artisans, from the end of the fourteenth century.13  In a most celebrated case, the book of drawings of Giovannino de’ Grassi, model animals seem to have been one of the most frequently copied items,14 and one could remark that fidelity in the copying of a drawing of a dog is less vital for, say, sailors than fidelity in the copying of a geographical map.  And of course the modes and practices of circulation, diffusion, and use, private or public, of these books of drawings have not yet been completely elucidated.  Several techniques for producing identical copies of a drawing were in use at the time—witness Cennini’s well-known account.15  Cennini also describes, in another chapter of his treatise, the techniques already in use at his time for the printing of images—on cloth and textiles.16  When Alberti concluded, around 1450, that his manuscript treatise on architecture should not be illustrated, printed images were already a relatively pervasive and multifarious component of daily life in most European cities.  We cannot blame Alberti and most of his contemporaries for failing to recognize in this marginal technology what would soon become a major agent of cultural and social change.  After all, not so long as ten years ago many intellectuals of our own day were persuaded that virtual images would remain confined to a niche market of home video-games, and that the Internet—witness the destiny of its unlucky French precursor, the Minitel—would basically develop into a global arena for electronic dating.

The destiny of the Internet will be better evaluated ten years from now.  The destiny of printed images in the Renaissance can now be evaluated with some historical distance.  To make a long story short, printed images soon crossed paths with the new invention of printing with moveable types; printed books were almost immediately illustrated with woodcut images, and fairly soon copperplate engravings were on the rise, too—either associated with printed books or as independent prints or albums: pixel rich copperplate engravings of architectural details were already on the market as early as 1515.17  As it has been recently pointed out, particularly by Myra Nan Rosenfeld, the publication in Venice of Serlio’s Fourth and Third Book in 1537 and 1540 marks the inauguration of the modern, printed and illustrated architectural manual—an invention that was to be successful, influential, and long-lived.18 

Serlio’s woodcut illustrations were printed together with his typographical text.  Serlio’s architectural discourse and his architectural drawings, as published in his treatise, are closely imbricated and deliberately complementary.  For the first time in history, a writer on architecture could safely refer his readers to his own images.  Similarly, for the first time in history, words and images could now compete on the same basis.  And since the outward and visible form of architecture can be much better described by a picture than by a thousand words, images soon got the upper hand to the detriment of traditional ekphrastic mediation.

This shift from words to pictures had a decisive effect on the whole process of architectural design.  Unlike their medieval predecessors, Renaissance architects could now imitate models that they had seen—and that, without embarking on any long journey: in most cases, a visit to a local bookshop would have been enough.  As a consequence, visual imitation became the object of a heated theoretical debate.  Finally, a new and easy method aimed at fostering and simplifying visual imitation of antiquity was elaborated.  This method was itself based on the awareness of the identical reproducibility of printed images.  Its keystone, the system of the five Renaissance orders, represents a watershed in the history of European architecture.  

Against this interpretation, which I have developed elsewhere,19 at least one fundamental objection must be raised.20  According to a well established and frequently cited theory, handmade architectural drawings were the main vectors for the transmission of architectural experience in the Middle Ages as well as in the Renaissance:21 if surviving evidence of medieval model books is scant, handmade architectural drawings, in general, and drawings of antiquities, in particular, were indeed largely circulated in the fifteenth and sixteenth centuries, copied and transmitted from a limited number of archetypes, sometimes even mass-produced in well-organized workshops; some scholars have suggested that these books of drawings should then be considered as the first and most important single factor for the formation and diffusion of Renaissance antiquarian lore and of Renaissance architectural theories—and this, before, during, and well after the rise of printed and illustrated architectural books and of printed architectural drawings.22 

The main argument that speaks in favour of this theory is chronology.  The interest of the humanists for visual documentation of antiquity antedates Gutenberg’s invention.  The time when mechanically reproduced images started to exert any noticeable cultural or social influence is not known for sure, and opinions on that matter diverge.  But the decisive turning point, the crossroad where antiquarian interest, mechanical images, and architectural theory meet, and generate a brand new and revolutionary architectural environment can be precisely situated at some point between 1515 and 1537.  Taking this chronology into account, some general considerations come to mind that might help to conciliate two apparently antagonistic interpretive theoriesObviously, difficulties in the reproduction of images would have discouraged the use of images only when reproduction was needed.  Technical architectural drawings, conceived as a means of communication between architect, patron, and building site, were not destined to unlimited reproducibility.  Sometimes even one original could have been enough; in other cases existing technologies (pouncing, tracing, stencils, templates, transparent paper, and the like) could guarantee the reliable reproduction of a limited number of identical copies.23  As mentioned before, Alberti could obviously tell the difference between documents that were destined to be reproduced (hence, they should use words) and documents that were not destined to be reproduced (hence, they could be drawn).  He chose different media for different messages.  Some existing corpora of Renaissance architectural drawings are likely to include technical drawings—drawings that were meant to be sent from the architect’s workshop to one or more of his building sites, associates, or patrons.  Witness Alberti’s double standard, this point-to-point transmission of architectural designs is unrelated to the problem of the communication of architectural knowledge to an undifferentiated public of unknown readers, far away in distant spaces and future times.  These disposable technical drawings could be produced more or less profusely—according to the organization of an architect’s workshop—precisely because they were not destined to reproducibility.  They were meant to be used once.  Their occasional transmission, and survival, happened by accident. 

Similar arguments can be applied to the well-known topic of architectural sketchbooks.  Architects and artists of all times have kept—many still keep—more or less portable notebooks (on tablets or paper or silicon chips) where they record memories, comments, and drawings of things seen or conceived.  Journaux intimes are occasionally meant for publication, but in most cases they are not.  Given the organization of Renaissance and late-medieval artistic workshops, some of these sketchbooks may in fact have been designed with an eye to a semi-private (or semi-public) circulation.  They might have been used collectively and freely within the same workshop, and sometimes they actually were a workshop product—a collective work supervised by one master.  How and to what extent these sketchbooks might have been circulated outside their original workshop is anyone’s guess, and no clear pattern emerges from the diverse and conflicting evidence at our disposal.  At some time between 1515 and 1517 Michelangelo seems to have borrowed (and copied) a book of architectural drawings from the workshop of Bernardo della Volpaia;24 but then, the Coner Codex is no ordinary book of drawings—actually it is a quite extraordinary item.  More than twenty years before, Michelangelo, then a teenager, had tried to borrow what must have been a more standard model book from the workshop of Domenico Ghirlandaio.  In vain: Ghirlandaio had kept his book for himself.  Out of envy, suggests Michelangelo’s biographer, Ascanio Condivi, who must have heard the story from Michelangelo himself.25  And in 1389 a painter was tried and sentenced for larceny in Poitiers, France, for stealing a book of drawings from the workshop of a colleague.26 

There is, nevertheless, ample evidence of the fact that some books of architectural drawings were indeed circulated, transmitted and copied, and sometimes copied again, in different locations, at different times, and by more than one artist.  Illuminated manuscripts had been handed down for centuries in this way, and in some cases the variance of their illustrations does not seem to have been uncontrollable.27  One might be reminded here of the most remarkable destiny of Giuliano da Sangallo’s personal sketchbook, now known as the Barberini Codex, a collection of architectural drawings that Sangallo had apparently started in 1465, at a rather precocious age.28  A part of this album was copied in another Codex, now known as Escurialensis, realized in Florence or Rome between 1490 and 1508, of which different parts were differently attributed to Domenico Ghirlandaio, to Giuliano da Sangallo himself, to Filippino Lippi, to Baccio d’Agnolo, or to one or more of the workshops of any of these.29  But the final version of the album was put together at the request of a customer, a Spanish nobleman, who, it has been suggested, on the point of leaving Rome to go back to his homeland, might have felt the urge to buy some illustrated souvenir of the Eternal City.  In fact, don Diego de Mendoza was not—or possibly not only—anticipating the standard behaviour of many tourists yet to be born.  As soon as he was back in his native Sierra Nevada, he used the rather diverse material collected in his Italian album to have a new castle built there in the Italian style.  Even more surprisingly, some pages of the album were actually sent back to Italy again, to Genoa, as models for stonecutters and sculptors who were to realize the decorations that were eventually shipped to Spain and assembled in the building.30  Here we have noticeable evidence of how architectural visual motifs could be passed on from paper (or parchment) to paper, and even from paper to stone, in this case crossing the Western Mediterranean three times over an extended period of time.  If the final results at La Calahorra castle were not very similar to the originals, as seen in Rome, one is nonetheless led to believe that visual conformity, albeit not attained, must in this case have been the purpose of the whole operation.

In the light of this evidence, it is worth reiterating that handmade copies and printed copies of the same drawing differ in one basic, essential, and in a sense ontological aspect.  A handmade copy, regardless of the disposition of the copyist and of his or her motivation to stick to the original or to change it, is always an act of creative imitation.  With few exceptions—that shall be dealt with presently—handmade copies are drawn without and outside the control of the author of the original drawing.  But if the drawing is printed, this condition is reversed.  Both the author and the public know that the print is an identical reproduction of its matrix.  The medium itself is the guarantee of the fidelity of the reproduction.  At the two ends of the chain of communication, the maker and the user of the print share this same persuasion: that the image is an identical copy of its matrix, and that the original matrix was destined to identical and theoretically unlimited reproducibility.  This mutual awareness affects the status of the image, its authority, its reliability, its trustworthiness, and in the end, the use that can be made of it. 

A growing need for precision in the transmission of visual information is the key factor in this process.  Of course printing does not guarantee the precision of the original drawing.  But it does guarantee to the author and to its audience that no middlemen will tamper with, or disrupt, the chain of transmission—from the maker to the user—of the image itself.  What the printer sees is identical with what every reader gets.  As a consequence, the general reliability of technical and scientific images was greatly enhanced.  The same was true for drawings of architecture, and to a more limited extent, and with different nuances, also for the other two arts of drawing.  In different domains, people found out that they could trust images as never before.  They also found out that they could buy more and more of them, at the same time as offer diversified, quality improved, and prices went down.  In short, this is a success story.  The direct consequence of this process was the rise of a new visual culture—a culture in which knowledge could be recorded and transmitted in a new visual format.

This success story does not imply that the availability of printed images was the primary cause for the rise of a new visual culture in the Renaissance, no more than it implies the opposite—that a new visual culture was the primary cause for the rise of printed images in the Renaissance.  In the case of such a sweeping cultural and technical change, relations of cause and effect can never be so clear cut.  Indeed, the invention of printed images can hardly be called an invention at all.  As has often been remarked, all that would have been technically necessary for printing images had always been available, from time immemorial—everything, except the idea that images should be printed.  The aspirations of a new visual culture, and the new technique that would eventually fulfil those same aspirations, must have come to maturity almost simultaneously.  Without the new visual needs of Renaissance culture, printed images might have existed—as they probably had existed for centuries—but they would never have spread, as they did in the fifteenth century.  Without printed images, the new visual needs of the Renaissance would soon have petered out.  As we know, they didn’t.  Starting from the end of the fourteenth century, and more particularly from the outset of Italian humanistic antiquarianism, proximity and interaction between a new graphic medium and the need for a new type of visual information created a favourable environment for the technical and cultural revolution that would eventually materialize early in the sixteenth century, when the decisive encounter between print and visual thinking unleashed a new and apparently unrestrainable wave of change.31 

In this unprecedented environment, where images could finally attain visual precision—and where the users of images had come to expect from images just that—it comes as no surprise that some handmade images could start, perhaps inadvertently, to imitate their more technological clones, or at least to pursue the same objectives.  The Coner Codex, already mentioned, features an amazing title page, where instructions are given on the units of measurement that were used in the restitution drawings, and the author personally guarantees that all the measurements in the book are precise.32  Unlike the author of a printed book, the author of the Coner Codex could not have extended his guarantee to any other copy than his own—the original one, the only one that he had himself licensed, signed, and sealed. 

However, there is proof that in some Renaissance workshops more identical copies of the same image were drawn, sometimes simultaneously, and sometimes by the same hand.  As a consequence, some scholars have suggested the hypothesis of a mass-production of handmade architectural drawings in the sixteenth century, drawings that would have been fabricated in well-organized workshops—the early-modern equivalent of medieval scriptoria, now no longer specialized in the production of alphabetic texts, nor of rare and rich illuminated manuscripts, but in the assembly and large scale reproduction of handmade, but standardized, architectural drawings.33  Existing evidence is too scarce to evaluate the scale of this phenomenon, and its possible effects.  One exception could be the workshop of the French Huguenot architect, businessman, printer, and wine merchant, Jacques Androuet du Cerceau (ca. 1520-ca. 1586).  It was recently pointed out that Du Cerceau’s house produced and sold printed, as well as handmade, copies of the same drawings.34  It seems that in some cases the handmade copies were not exactly identical, and that deliberate and substantial variances were introduced in some drawings; in other cases, the manual copies seem ostensibly identical to each other, and also identical to the printed version.35  Further investigations should elucidate to what extent the publisher’s choice between print and drawing might have been related to the subject of the image, to the public that was being targeted, or both;  which kind of variants were introduced in the handmade items, and why.  A careful collation between the printed and the manual version of the same drawing should consider the mutual influences between the two media, and might be able to tell which of the two was setting the trend.  For example, while it is evident that early prints imitated the graphic style of drawings, it may also be inferred that some handmade drawings in the sixteenth century tried to emulate the graphic style of prints.36 

No data are available on the pricing policy of the Du Cerceau publishing house, and no comparison is possible between the number of copies, either printed or handmade, that were being produced of the same item.  For all we know, the phenomenon of handmade, customized drawings might have been limited to a very small number of de luxe copies, in the same way as handwritten and printed copies of the same book coexisted for a time well after the rise and the universal diffusion of the printing press.  They coexisted, but they did not really compete: it is known that in Florence, in 1483, the price of a manuscript copy of Ficino’s translation of Plato’s Dialogues must have cost 300 to 400 times the price of the printed version of the same.37  If such figures were to be confirmed with regard to illustrated books and printed drawings in the sixteenth century, it would be easy to conclude that handmade architectural drawings must have occupied a marginal, upper end, and, in fact, ephemeral segment of a market that in the sixteenth century was already driven and dominated by their mechanical equivalent.  As suggested by Arnold Nesselrath, handmade architectural drawings, and drawings from the antique, would soon cease to be considered as viable means for the transmission of factual information; instead, at some point in time—probably early in the seventeenth century—they received a new lease on life as self-standing and autonomous artistic expressions.38 

Pending the digging out of new facts and figures on the matter, it seems nevertheless safe to conclude that the reproduction of identical manual drawings in the sixteenth century should be seen in the context of the new visual environment that was being created by the rise of mechanically reproduced images.  The diffusion of printed images had fostered new and higher standards of precision in the transmission of visual knowledge.  People expected graphic documents to be accurate and reliable: reproduced drawings had to be identical to their original, and certified as such.  In this context, evidence to be further investigated would seem to suggest that the sign of the print, rather than the signature of any individual artist on his or her handmade drawings, soon came to be considered as the trademark of documentary trustworthiness.  In retrospect, this should come as no surprise: modern history has proved, more than once since the first technologizing of the printed page in the Renaissance, that when we aim at identical reproduction, machines can generally deliver it better than we can. 

footnotes

* A preliminary version of this article was presented at the fourth meeting of the International Society for the Classical Tradition, held in Tübingen from July 29 to August 2, 1998.  I am indebted to Professor Wolfgang G. Haase for numerous comments and for his careful reading of the original draft, and to Myra Nan Rosenfeld, who kindly provided many useful data and references (further acknowledged in the footnotes).

1. Vasari, Vite, (15501, 15682), in  Le opere di Giorgio Vasari, con nuove note e commenti di Gaetano Milanesi, Florence: Sansoni, 1878-85, vol. II, 1878, 339-40. 

2. Vasari’s passage is cited and discussed, with regard to the early modern transition from orality to visual media, by Arnold Nesselrath, «I libri di disegni di antichità: tentativo di una tipologia», in  Salvatore Settis, ed., Memoria dell’antico nell’arte italiana, vol. III, Dalla tradizione all’archeologia, Turin: Einaudi, 1986, 89-153: 99 and footnotes 28, 29 (“Qui manifestamente alla descrizione verbale viene contrapposta quella visuale, effettuata col mezzo figurativo”).

3.  See Mario Carpo, L’architettura dell’età della stampa. Oralità, scrittura, libro stampato e riproduzione meccanica dell’immagine nella storia delle teorie architettoniche, Milan: Jaca Book, 1998;  Id., «The Making of the Typographical Architect», in Vaughan Hart, Peter Hicks, eds., Paper Palaces. The Rise of the Renaissance Architectural Treatise, New Haven, CT: Yale University Press, 1998, 158-170.

4. Julius von Schlosser, «Zur Kenntnis der künstlerischen Ueberlieferung in späten Mittelalter», in  Jahrbuch der kunsthistorischen Sammlungen des allerhöchsten Kaiserhauses, xix, 1902, 279-286, 318-326.  For a survey of the current state of research on this celebrated question see Robert W. Scheller, Exemplum. Model Book Drawings and the Practice of Artistic Transmission in the Middle-Ages, ca. 900 - ca. 1470, Amsterdam: Amsterdam University Press, 1995, 7-10, 48-87.  See also Ernst Kitzinger, «The Role of Miniature Painting in Mural Decoration», in  Kurt Weitmann, ed., The Place of Book Illumination in Byzantine Art, Princeton: Art Museum, Princeton University, and Princeton University Press, 1975, 99-143;  Kurt Weitzmann, Illustrations in Roll and Codex.  A Study of the Origin and Method of Text Illustration, Princeton: Princeton University Press, 1947 (revised ed., Princeton: Princeton University Press, 19702);  Id., Ancient Book Illumination, Cambridge, MA: Harvard University Press, 1959, chapter I, “Scientific and Didactic Treatises”, 5-31.

5. Richard Krautheimer, «Introduction to an ‘Iconography of Medieval Architecture’», in Journal of the Warburg and Courtauld Institutes, v, 1942, 1-33; revised ed. in  Id., Studies in Early Christian, Medieval, and Renaissance Art, New York: New York University Press, 1969, 115-150.

6. Richard Krautheimer and Trude Krautheimer-Hess, Lorenzo Ghiberti, Princeton: Princeton University Press, 1956, 294 (“To Petrarch […] it mattered little whether or not a site was commemorated by a monument, or merely haunted by memories.  His approach was entirely literary, almost emphatically nonvisual.”)  This passage is cited and discussed in a similar context by Françoise Choay, L’allégorie du patrimoine, Paris: Seuil, 1992, 39 and footnote 31.

7. Walter Ong, Orality and Literacy: The Technologizing of the Word, (1982), London and New York: Routledge, 19955.  See in particular chapter iv, “Writing is a Technology”, 81-83.

8. See William M. Ivins, Prints and Visual Communication, (1953), Cambridge, MA, and London: The MIT Press, 1978.  Ivins, who minted the formula “exactly repeatable pictorial statements” (23), discussed the far reaching implications of this notion for the history of science, technique, philosophy, and the fine arts.

1. Carpo, «The Making of the Typographical Architect», 1998, 163-64 and footnote 5, with further bibliography.

10. Id., L’architettura dell’età della stampa, 1998, 155-56, with further bibliography.  Some decades after the rise of a new visual environment based on ubiquitous and reliable printed images, Pliny’s suppression of illustrations was bitterly lamented by Cardinal Federigo Borromeo (“Moreover, how much light would we glean from interpreting the passages of writers, principally Pliny, if we had in sight those things that he told only with words”: Federigo Borromeo, Musaeum, 1625; see Arlene Quint, Cardinal Federigo Borromeo as a Patron and Critic of the Arts and his Museum of 1625, New York: Garland, 1986, 233; Paula Findlen, «The Museum: Its Classical Etymology and Renaissance Genealogy», in Journal of the History of Collections, 1, 1989(1), 59-78: 65, and 75, note 46).  Apparently, in 1625 Pliny’s arguments against the manuscript transmission of scientific drawings already fell on deaf ears.

11. Ivi, 156-57;  Mario Carpo, «Descriptio Urbis Romae.  Ekfrasis geografica e cultura visuale all’alba della rivoluzione tipografica», in  Albertiana, (Florence: Leo S. Olschki Editore), i, 1, 1998, 111-142: 126-134, with further bibliography.  (See now also Leon Battista Alberti, Descriptio Urbis Romae. édition critique, traduction et commentaire par Martine Furno et Mario Carpo, Geneva: Droz, 2000, 64-96).  

12. On Alberti’s explicit and reiterated refusal to illustrate his manuscript treatise on architecture, De re aedificatoria, see Françoise Choay, La règle et le modèle. Sur la théorie de l’architecture et de l’urbanisme, (1980), Paris: Seuil, 19962, 136-140 (English transl., ed. D. Bratton, Cambridge, MA: The MIT Press, 1997);  Ead., «Le De Re Aedificatoria comme texte inaugural», in Jean Guillaume, ed., Les traités d’architecture de la Renaissance, actes du colloque de Tours, 1-11 juillet 1981, Paris: Picard, 1988: 85-90;  Ead., La regola e il modello, revised transl., transl. E. d’Alfonso, Rome: Officina, 1986, 141, footnote 134;  M. Carpo, Metodo e ordini nella teoria architettonica dei primi moderni, Geneva: Droz, 1993, 13-24;  Id., L’architettura dell’età della stampa, 1998, 127-132.  Even more strikingly, and presumably for the same reasons, Alberti adopted a system of polar coordinates to convert into a string of letters and numbers the drawing of a map of Rome that he had himself surveyed, carefully measured, and drawn to scale: see Carpo, Descriptio Urbis Romae, 1998 (and Alberti, 2000: see note above).

13. Scheller, 1995, 1-7, with further bibliography.

14. Circa 1380-98.  See Scheller, 1995, 287.

15. Cennino Cennini, Il libro dell’Arte, (circa 1400), eds. Carlo and Gaetano Milanesi, Firenze: Le Monnier, 1859, chapters 23-26.  Passages cited and discussed in Scheller, 1995, 38, 72-74 and footnote 203.

16. Cennini, 1859, chapter 173: see Scheller, 1995, 75;  Arthur Mayger Hind, An Introduction to a History of Woodcut, With a Detailed Survey of Work Done in the Fifteenth Century, London: Constable and Co., 1935, vol. I, 4-6. 

17. See A.M. Hind, Early Italian Engraving. A critical catalogue with complete reproductions of all the prints described. London: Bernard Quaritch, 1938-48: II, V, (1948), 52-53, 268-269;  Nesselrath, 1986, 126-27, footnote 18 with further bibliography.

18. Myra Nan Rosenfeld, «Sebastiano Serlio’s Contributions to the Creation of the Modern Illustrated Architectural Manual», in Christof Thoenes, ed., Sebastiano Serlio, Atti del Convegno, Vicenza 31 agosto - 4 settembre 1987, Milan: Electa, 1989, 102-110.  See also, Ead., Sebastiano Serlio on Domestic Architecture: Different Dwellings From the Meanest Hovel to the Most Ornate Palace. The Sixteenth-Century Manuscript of Book VI in the Avery Library of Columbia University, New York: The Architectural History Foundation, and Cambridge, MA: The MIT Press, 1978, 37-41;  Ead., Serlio on Domestic Architecture, Mineola, New York: Dover Paperback, 1996, 1-8.

19. Carpo, L’architettura dell’età della stampa, 1998;  Id., «The Making of the Typographical Architect», 1998.

20. And, indeed, this objection has been raised, and I am grateful in particular to Myra Rosenfeld, Frédérique Lemerle, and Yves Pauwels for bringing the topic to my attention.  See Frédérique Lemerle, review of Carpo, L’Architettura dell’età della stampa, in Revue de l’art, 124, 1999 (2): 85-86; and Yves Pauwels, another review of the same, forthcoming in the Journal of the Society of Architectural Historians [59, 2000 (3)].

21. On the transmission of artistic ideas in the Middle Ages, see note 4, above.  An equally vast bibliography is available on Renaissance model books and sketchbooks, and also more specifically on the relationship between drawn and printed model books in the fifteenth and in the sixteenth century: for the most recent contributions, see Konrad Oberhuber, «Introduction», in  Jay A. Levinson, Konrad Oberhuber, Jacquelyn L. Sheehan, Early Italian Engravings from the National Gallery of Art, Washington: The National Gallery of Art, 1973, xiii-xxvi;  Joseph George Rushton, Italian Renaissance Figurative Sketch-books, 1450-1520, Ph.D. Dissertation, 1976, Ann Arbor: University Microfilm International, 1981;  Hubertus Günther, Das Studium der antiken Architektur in den Zeichnungen der Hochrenaissance, Tübingen: Ernst Wasmuth, (Veröffentlichungen der Bibliotheca Hertziana in Rom), 1988;  Arnold Nesselrath, 1986;  Id., «Il codice Escurialense», in Wolfram Prinz, Max Seidel, eds., Domenico Ghirlandaio, 1449-1494, Atti del Convegno, Firenze 1994, Florence: Centro Di, 1996, 175-199;  Id., «Das Liller ‘Michelangelo-Skizzenbuch’», in Kunstchronik, 36, 1983, 46-47;  Id., «Il libro di Michelangelo a Lille», in Quaderni dell’Istituto di Storia dell’Architettura, (Rome: Bonsignori), n.s., 24, 1997, 283-322;  Myra Nan Rosenfeld, «From Drawn to Printed Model Book: Jacques Androuet du Cerceau and the Transmission of Ideas from Designer to Patron, Master Mason and Architect in the Renaissance», in RACAR, xvi, 2, 1989, 131-147;  Frédérique Lemerle-Pauwels, «Le Codex italien du Musée des Beaux-Arts de Lille: les modèles d’architecture antique et moderne de Raffaello da Montelupo, 1504-1566», in Revue du Louvre et des Musées de France, 1997, 2, 47-57;  Ead., «Le livre de dessins de Michel-Ange», in Catalogue des dessins italiens. Collection du Palais des Beaux-Arts de Lille, Paris and Lille: Réunion des Musées Nationaux and Palais des Beaux-Arts de Lille, 1997, 283-322;  R. Caltarossa, «Il codice di Oreste Vannocci Biringucci nel contesto dei codici del Rinascimento», in Annali di Architettura, 8, 1997, 43-60.

22. This thesis was and still is to some extent taken for granted by several art historians, and it has more recently been clearly outlined in Nesselrath’s essay, «I libri di disegni di antichità», cit., 1986.  According to Nesselrath, the printing of images and of illustrated treatises could seldom compete with the copying and circulation of handmade drawings (artists’ sketchbooks and drawn model books) throughout most of the sixteenth century, for reasons that will be discussed hereafter; only towards the end of the century did the growing diffusion of mechanically reproduced images bring about a clear differentiation of media and contents: at a given point in time, artists’ drawings came to be considered as an “autonomous artistic expression” (ivi, 144), but this turning point “marks the transition from the Renaissance to Mannerism and the Baroque, more than the transition from the Middle Ages to the Renaissance” (ivi, 146, translation mine).

23. See above, note 15;  Carpo, L’architettura dell’età della stampa, 1998, 176-77, with further bibliography.  Nonetheless, some technical drawings specifically meant for the building site are also known to have been printed.  This is for example the case with a number of project drawings for the church of the Invalides in Paris, reproduced from copperplate engravings in order to be attached to the construction contracts of December 7, 1690.  The copies in possession of the Centre Canadien d’Architecture, Montréal, are published and discussed in L’Architecture et son image. Quatre siècles de représentation architecturale, œuvres tirées des collections du CCA, Edited by Eve Blau and Edward Kaufman, Montréal, CCA, 7 mai- 7 août, 1989, 164-165, ill. 166-167, 5.1-5.7.  See also Patrick Reuteswärd, The Two Churches of the Hôtel des Invalides, a history of their design, Stockholm: Nation-museum, 1965, 89-91, 27, and 90, note 107. One may infer that many identical drawings were needed here because these sorts of construction contracts had to be, in a sense, published in many identical copies, hence prints, rather than drawings, would help. (This case and the related sources were pointed out to me by Myra Rosenfeld).  In more recent times, blueprints, or heliographed copies destined for the building site did not actually rank as printed drawings, and as a rule were not considered as such.  Today, computer generated images have created a new hybrid medium, which blurs the traditional borderlines between individual, handmade drawings and reproducible but unalterable clichés.  In this context, the semantic drift of the very same words “cliché”, “stereotype”, “bromide”, and the like, all of them originally technical terms related to the reproduction of identical texts or images (or, in the case of “qwerty”, to typing), deserves a closer socio-linguistical investigation.  A computational analysis would show that “bromide”, now obsolete, is one of the (negative) keywords in Ayn Rand’s notorious novel on architecture, The Fountainhead (1943).

24. A group of drawings of Roman antiquities and architectural details, dated 1515-17, attributed to Michelangelo or one of his assistants, and now divided between the Casa Buonarroti in Florence and the British Museum in London, are copies from the Coner Codex or from a source common to both.  These were first identified by Ashby, 1904: see Giulio Carlo Argan, Bruno Contardi, Michelangelo architetto, Milan: Electa, 1990 (French transl., D.-A. Canal, Milan and Paris: Gallimard-Electa, 1991, 154); Nesselrath, 1986, 136, footnote 12, with further bibliography.

25. Nesselrath, 1986, 129, and footnotes 8, 9, with further bibliography.  Nesselrath suggests that this book of drawings could not have been the Codex Escurialensis, which some authors, starting with H. Egger, 1905-06, have attributed to Domenico Ghirlandaio’s workshop (against this attribution, see Id., ivi, 130-134; Id., 1996, 175-99).

26. Alfred de Champeneaux, Paul Gauchery, Les travaux d’art exécutés pour Jean de France, duc de Berry, avec une étude biographique sur les artistes employés par ce prince, Paris: H. Champion, 1984, 205; passage cited and discussed in Scheller, 1995, 79 and footnote 217.  

27. Ample evidence of that is given and discussed by Weitzmann, 1947, 1959, and 1970 (works cited above, note 4).

28. On the Barberini Codex see Nesselrath, 1986, 127-29, with further bibliography.  The date 1465 features on the title page of the Codex, probably a later addition since the author uses there his toponym “da Sangallo” (after the name of a district in Florence), which he seems to have adopted only around 1483.  The date of birth of Giuliano da Sangallo is not known for certain (perhaps 1443 or 1445);  Nesselrath suggests that in 1465, when he apparently started collecting his own drawings, Sangallo would have been 13 years old.  The Codex was inherited by Sangallo’s son, as a private and personal belonging of his father’s (ivi, 129).  Nevertheless, scholars have identified scores of direct or “parallel” copies of Sangallo’s drawings in several other Renaissance books of drawings (most notably in the Codex Escurialensis: see hereafter): Nesselrath, 1986, 130-31, and Id., 1996, 189-90, with further bibliography.

29. Hülsen proved in 1910 that the third part of the Codex Escurialensis is a direct copy of the oldest part, called “Libro piccolo”, of Giuliano da Sangallo’s Codex Barberini: Nesselrath, 1986, 130-31, and Id., 1996, 187-89.

30. On the Codex Escurialensis, see Nesselrath, 1986, 129-134, and Id., 1996.  Don Diego (or Rodrigo) de Mendoza marked with a cross the models in the book that he wanted executed for his Spanish castle, and to this end he seems to have sent his own original copy of the Codex to the sculptors in Genoa: see Nesselrath, 1986, 134, footnote 21, and Id., 1996, 197, footnote 25, with further bibliography;  Fernando Marias, «Sobre el Castillo de la Calahorra y el Codex Escurialensis», in Saggi in onore di Renato Bonelli, I, (Quaderni dell’Istituto di Storia dell’Architettura, xv-xx, 1990-92), Rome, 1992, 539-53.

31. Although focusing on visual rather than on textual communication, some of these general arguments are inspired by, and heavily indebted to, the works of Elizabeth L. Eisenstein, The Printing Press as an Agent of Change: Communications and Cultural Transformations in Early Modern Europe, London and New York: Cambridge University Press, 1979, and The Printing Revolution in Early Modern Europe, Cambridge (U.K.): Cambridge University Press, 1983. 

32. “Nota quod omnia quae in isto libro sunt menssurata cum brachiis florentinis dividendo brachium in partes sexaginta quas voco minuta et cum ipsis minutis minutissime mensuratum est” (Bernardo della Volpaia, Coner Codex, f. 1r.: London, Sir John Soane Museum.  Reproduced in Nesselrath, 1986, fig. 121).  On the Coner Codex and its many copies, including Michelangelo’s, see Nesselrath, 1986, 136-37, footnotes 12-13, with further bibliography.  Nesselrath discusses Bernardo della Volpaia’s (and other contemporary artists’) quest for precision, and he concludes that these requirements could not be met by printed pictures, either xylographed or engraved, but only by handmade drawings, with the possible exception of Serlio’s influential woodcut illustrations in 1537 and 1540 (ivi, 138-39).  Nesselrath does not take into account the topic of reproduction, and this flaws some of his arguments: admittedly, one drawing on parchment or paper can be more precise than the same drawing when cut in a wood-block or even engraved on a metal plate; but, for better or worse, a  mechanical matrix can generate many identical copies—a drawing on paper cannot.  It may be argued that, in quantitative terms, the amount of fidelity to the original drawing that is lost in the transfer of images from paper or parchment to wood or metal plate is more than compensated by the fidelity to the mechanical archetype that is gained through the identical and unlimited reproduction of it.  Failure to consider these factors, and their role in sixteenth century artistic culture, may tilt the balance when assessing the close relationship that existed between drawn and printed model books during the Renaissance.  Likewise, these premises are likely to propitiate, perhaps inadvertently, some factual inaccuracies: for example, Nesselrath emphasizes the close complementarity between text and images in Vitruvius’s and Alberti’s manuscript treatises (ivi, 135), and he refers to a seamless unity of text and illustrations as Alberti’s goal, albeit one that Alberti failed to achieve (ivi, 106).  The cases of Francesco di Giorgio’s and Filarete’s manuscript treatises, both illustrated, may be misleading: I have discussed them elsewhere (Carpo, L’architettura dell’età della stampa, 1998, 132-150).

33. Evidence cited by Nesselrath includes the sketchbook of Raffaello da Montelupo now at Lille, and another identical book of the same author, of which only a fragment survives (Rugby School); the two identical codices of the so-called Master C of 1519 now at Chatsworth and at the Albertina; the two almost identical codices, drawn by different copyists, known as the Destailleur B (Berlin) and Destailleur (Vienna); also, two different but comparable cases, Giovanni Battista Montano’s simultaneous production of handmade, and printed, versions of the same drawings; and the workshop of Du Cerceau, of which hereafter: Nesselrath, 1986, 135-40, with further bibliography.  Likewise, Myra Nan Rosenfeld has discussed a drawn model book by an anonymous Italian artist (circa 1520-30), now at the Centre Canadien d’Architecture.  All but eleven of the twenty-two drawings are identical to those found on a sheet in the Uffizi (A 689), formerly ascribed to Sallustio Peruzzi, and f. 23r., the depiction of the Licinian Garden Pavilion, was copied from the drawing in the Codex Coner on f. 15r. (see Rosenfeld, 1989, 138 and note 47). 

34. On Du Cerceau’s diverse publishing activities see Janet Byrne (1977, et al.) and David Thomson (1988), cited and discussed by Myra Nan Rosenfeld, 1989, 133 (“Du Cerceau directed a large workshop of designers, draughtsmen, and printers […]. As Byrne has noted, Du Cerceau must have provided the original designs and had the members of his workshop execute the books, drawings, and prints, in a mass-production fashion”).  Based on the study of two suites of engravings by Du Cerceau, and of several collections of drawings attributed to Du Cerceau’s workshop now in the collections of the CCA, Rosenfeld has concluded that Du Cerceau must have frequently used Italian drawn model books as a source for his own publications (ivi, 137-38), and that Du Cerceau’s workshop produced both drawn and engraved versions of the same design (ivi, 138).  But see D. Thomson, «Jacques Androuet du Cerceau. An Album of Forty Architectural Drawings», Old Master Drawings, London: Sotheby’s, March 25, 1982.  Bruno Adorni, «Il problema dei disegni della bottega di Du Cerceau», in Il disegno di architettura, iv, 1993, 8, 10-17, does not provide new information.

35. Rosenfeld, 1989, 138-39, footnote 49, 136, and passim.  I refer here to oral communications from David Thomson and Myra Nan Rosenfeld.  Rosenfeld suggests that it is possible to identify several different masters in Du Cerceau’s workshop, according to support used (vellum or paper), type of ink, and drawing technique.  This subject will be discussed in her forthcoming book: From Model Book to Printed Treatise: Creating a Common Body of Knowledge about Architecture, 1400-1600.

36. Myra Rosenfeld has identified several well-known early sixteenth-century architectural drawings that were drawn to look like prints.  In a couple of cases, the fake is so convincing that it has indeed misled some eminent scholars.  This subject will be further discussed in Rosenfeld’s forthcoming publication (see note above).

37. Eisenstein, 1983, 17, with further bibliography.

38. Nesselrath, 1986, 144-46.  See above, note 22.