For the last three years paper conservator Angela Campbell and artist-engraver Andrew Raftery have been engaged in an innovative research project aimed at answering certain questions about Albrecht Dürer’s working methods and about the physical life of engraved plates—the sources of some of the most powerful and influential images of the 16th century. Below, they talk about their collaboration.
Albrecht Dürer’s engravings, particularly his three Meisterstiche (“master engravings”) Knight, Death, and the Devil, Melencolia I, and St. Jerome in His Study, have been studied, copied, and adored for the better part of 500 years. Almost every institutional and many private print collections in the US and Europe have an impression of at least one of his three best-known prints. And while many of these impressions are meticulously labeled with descriptive adjectives, determining their comparative quality can be enormously challenging. Over the course of a two-year Andrew W. Mellon Fellowship in the Sherman Fairchild Center for Works of Art on Paper and Photograph Conservation at the Metropolitan Museum of Art, and with the indefatigable help of Andrew Raftery, engraver extraordinaire, I endeavored to provide and codify a method for quantitatively comparing Meisterstiche impressions by determining their comparative chronological place.1
These three works were, literally, Dürer’s master engravings. Dürer knew this, even as he was engraving the plates, and the hundreds of Meisterstiche impressions that remain in existence today confirm that even the earliest collectors to acquire them were aware of their extraordinary nature and value; for so many impressions to remain accessible today, they would have had to have been coveted and well-protected over the tumultuous course of centuries.
These surviving impressions, not surprisingly, have been the subject of endless comparison: every print custodian wants to know that his/her impression is among the very best and/or the earliest (this is sometimes thought to mean the same thing, though it often does not). In 1932, Joseph Meder published a detailed text2 that attempted to categorize all of Dürer’s prints by state and sub-state. Meder’s text, like other similar works, is limited by its reliance on describing the different Meisterstiche impressions subjectively with words such as ‘rich,’ ‘clean,’ ‘with burr,’ ‘excellent’ or ‘poor,’ ‘lighter’ or ‘darker,’ ‘silvery’ or ‘brownish.’ Curators have striven to illustrate these qualitative differences through exhibitions that include multiple impressions of the same engraving, but the determination of a good/better/best impression will always remain subjective. Determining the chronology of impressions, on the other hand, could potentially be an objective exercise from which information about a particular impression or collection of impressions may be gathered and used to inform later studies. Meder understood this, but writing in the 1930s, without access to the photodocumentation tools available today, there was little he could do.
A portable digital photography system capable of high-resolution, magnified reproduction was developed for this study.3 By comparing the images captured with this system, details of plate-wear visible in the individual impressions could be analyzed and an impression chronology could be built. Close-ups were taken of a small section of the head of the Knight’s steed in three impressions of Knight, Death and the Devil (1513), two in the collection of The Metropolitan Museum of Art, and one from the Frick Collection.
Fine scratches in the plate, such as the horizontal scratch clearly visible in the the first Met impression, are present in various areas of all Meisterstiche impressions, whether they are early, middle or late impressions. With repeated printing, lightly engraved lines wear away more quickly than deeply engraved lines, and these fine scratches wear away still more rapidly, so can act as indicators of chronological order. Comparing these enlargements, it is clear that the highlighted scratch disappears almost entirely. In this case, the scratches only confirm what is visible in the overall impressions—the first impression is early, the second impression was printed sometime in the middle of the plate’s life, and the third impression is late. The subtleties in building the chronology lie between these impressions: using this specific scratch and comparing it to the appearance and fading of other scratches scattered throughout the impressions, a much more specific order of printed impressions can be constructed. The compiling of chronological information continues, and I hope eventually to make the material available to scholars through what will likely be an online database.
In addition to chronological information about the impressions, the images revealed further evidence about the manufacturing methods used to produce Dürer’s plates and inks, as well as the artist’s meticulous selection of both engraving tools and madeto-order paper. It was with this information that I approached Andrew Raftery and asked for his assistance in creating a copy of one of Dürer’s engravings with as much historical accuracy as possible. We hoped this would produce a deeper understanding of how Dürer engraved his plates and how those plates wore down over time, and would also help test the validity of my chronology-building method.
I first met Angela Campbell when she was researching Melencolia I. Her close examination of 140 Meisterstiche impressions (59 of Melencolia I) and her extensive knowledge of the physical properties of materials had led her to think about the printed impression’s unseen corollary, the copperplate. Angela approached me about making a plate that could be used to test some hypotheses about how Dürer had worked and how the plates had aged.
At the time I happened to be engraving a copy after Cornelis Cort’s 1573 St. Dominic, based on a design by Bartholomeus Spranger. It may seem an odd thing for a contemporary artist to do, but copying is an age-old practice. At its worst, it is simple mechanical imitation (or even deception, in the case of forgery), but it is only through copying that an artist can internalize the formal qualities, visual language and techniques of another artist.
Dürer’s copies of Mantegna engravings, for example, explored the Italian artist’s low relief and classicizing form through Dürer’s own florid calligraphic pen work. Marcantonio’s engraved copies after Dürer’s woodcuts may have been motivated by the market, but we can see how his careful replication of Dürer’s marks led him to a new understanding of crosshatching that he eventually synthesized in his engravings after Raphael, which in turn were widely copied by other artists.
Coming of age in the 1980s, at the height of appropriation art, I always thought traditional copying had fresh possibilities that were too little discussed. My own feeling after copying a print or drawing is that I now own something about it that can be incorporated into my own work. When Angela approached me, I was curious whether a contemporary artist could use copying to create new knowledge of value to art history.
The Cort copy, I realized, would have limited relevance to the research on Dürer: Cort’s deep swelling lines are very different from Dürer’s wiry threads. I knew from experience that copying Dürer would be much more daunting. In 2009, while working on the “Brilliant Line” exhibition and catalogue,4 I had made a series of layered drawings on clear acetate that analyzed systems of line used by various 15th through 17th century engravers. One of the works studied was Dürer’s 1511 engraving Madonna with the Pear, which was by far the most difficult. His hatching strategies are sublimely effective at articulating form, but they defied the predictable logic that characterized every other engraver in my study. As uncomfortable as I am with the idea of genius, such close study of Dürer reaffirms what we already know: he was amazing.
For the proposed copy, Angela and I selected Dürer’s St Paul (1514), which had been engraved the same year as Melencolia I. A small print showing a single, draped figure before a landscape, it was especially interesting to me because there is in Vienna a preparatory pen drawing of the figure (in reverse) in which we can see Dürer working out the hatching of the drapery, and the Berlin Kupferstichkabinett has an early state of the engraving which demonstrates that the figure was completed before the background. The Met has a fine impression of the final state and Angela made a high resolution scan for me to study as I engraved.
Although it would be impossible to replicate perfectly Dürer’s materials or processes, my aim was to consider how an engraving might have been made in the early 16th century, starting with the raw plate. The copperplates available today are rolled in manufacture, which creates a pronounced grain in one direction that influences the action of the burin. (More resistance is felt while going against the grain while the burin wants to move quickly in the direction of the grain.) Until well into the 20th century, however, plates were made by hammering, which resulted in an evenly hardened surface.
Working with Brian Bergeron,5 technician in RISD’s Department of Jewelry and Metalsmithing, we set out to make a hammered plate: we liquefied the copper using an acetylene torch (it took a very long time), cast a ¼ inch thick ingot, hammered it on an anvil until the “ping” of the hammer changed tone (an indication that the copper had been compressed as much as possible), annealed the copper by heating it until it was red hot then plunging it into cold water, and then repeated the hammering and annealing cycle. Unfortunately, though the ingot had looked solid when first cast, hammering revealed a crack that would only widen as the plate got thinner.
Brian suggested that instead of an ingot, we start with a thick rolled plate since “the copper won’t know” how it had been formed prior to hammering. A sixteen-gauge plate was hammered, annealed, hammered, annealed and given a final hammering. The plate, which started at 7 ¼ x 4 inches, grew ¼ inch in each direction, and thinned to nineteen gauge. The whole process gave me new respect for our Bronze Age ancestors, not to mention 16th-century coppersmiths. I now understand why they did those eight-year apprenticeships.
After hammering I smoothed and polished the plate with an orbital sander. The final result was a revelation: The surface had a glassy quality unlike any plate I had ever used, and under the burin it was both supple and springy. Angela and I had decided that I should use a lozenge-shaped burin for the entire image as it allows greater depth on fine lines that the square burins I normally use. Greater depth means longer wear for the plate and we thought the fine, clear lines would be appropriate for Dürer. On modern rolled copper, the lozenge throws up rough edges when engraving curved lines against the grain. The hammered copper made me understand why the lozenge tool even existed. The plate and tool had unlimited potential.
The design was transferred onto the plate by taking a reversed photocopy of the Dürer St. Paul, placing it on the wax-coated plate and tracing over the outlines with a stylus. The outlines impressed into the wax were then reinforced with a steel drypoint tool. After the wax was removed from the plate, the burr was polished from the drypoint lines with a fine abrasive pad, leaving faint incised contours that could be followed throughout the engraving process.
I began engraving with the drapery, moving from the hem through the mantle to the sleeves. I engraved the outlines first and then the various layers of hatching, completing one area at a time. From my experience on the Madonna with the Pear I knew the hatching structure of the drapery would be complex and unpredictable. Although Dürer’s drapery forms relate to Gothic painting and sculpture, his resolution of volume is much more complete than that found in the work of other Northern Gothic engravers such as Schongauer. Dürer seems to improvise as he lays down patches of hatching and then responds to them with opposing lines that complete the form. It took all my concentration to follow Dürer’s lines. I regret to say that even with magnification and the highest level of mental focus I could muster, I was never able to fit quite as many lines in any given area as existed in the original. Dürer’s finesse eluded me. At some level it was impossible to move beyond my own hand as an artist. Perhaps this is the greatest frustration of the copyist.
And the drapery was the easy part. No engraver has ever matched Dürer’s depiction of hair, and the beard of St. Paul is an exquisite cascade that never loses its sense of underlying volume. I had to accept that my copy would be at best a humble tribute.
It took me just under 40 hours to engrave the plate and sharpen the tools. The St. Paul was one of eight engravings Dürer completed in 1514, including the large Melencolia I and St. Jerome in his Study. Keeping in mind Dürer’s virtuosity and the fact that he was the inventor of the St. Paul, it is plausible that he completed the plate in less time than it took me to do my laborious copy.
The day of reckoning came when I did a workshop for the paper conservation staff and the print curators at the Met. Their beautiful impression of St. Paul was on an easel. I placed my copy next to it. It was cruder but it displayed, I felt, a vitality that made it feel like a work of art in its own right. On the other side was an extraordinary copy of St. Paul by the Flemish fine-manner engraver Johannes Wierix. Completed when the copyist was only 17 years old, it was actually engraved with finer lines than the Dürer, using a delicate, even decorative calligraphy in the organization of the lines, that betrayed yet another set of artistic sensibilities in the hand of a copyist.
Making the plate is just the start of our research. The intention is to print the plate until it wears out, using the same ink on the same paper and the same press, with consistent wiping techniques and pressure, for the entire run. The hope is that such an experiment will give some idea of how many impressions a 16th-century plate might have yielded. So far we have printed well over 100 impressions. Not surprisingly, the 100th impression is fine, but significantly lighter than the 10th. The burr thrown up by the burin was polished off as I finished engraving each area, but enough remained on the plate to give about twenty exceptionally rich impressions. By the 50th impression the plate had stabilized, giving consistently clear, strong impressions. It will take much more printing, perhaps up to 2000 impressions, before we can come to any definitive conclusions about the lifespan of the plate.
There is, of course, much guesswork here. We do not know exactly how deeply Dürer’s plates were engraved to begin with nor what shape tool he used. Angela has recently seen the only surviving Dürer plate, the Portrait of Philip Melanchthon (1526), now in Gotha, Germany. She described the plate as very delicately engraved, but how much of that delicacy is due to wear is impossible to determine.
Impressions of my St. Paul are printed with modern ink, Gamblin Stiff Portland Black. How different is this ink from those used in the 16th century and how does the structure of contemporary pigments contribute to wear on the plate? The paper selected for printing, Ruscombe Mills Queen Anne Pale Laid, feels like an antique paper. It has a high linen content and is well sized. To what extent does the paper bear on plate degradation? What cloths were used to wipe the plates 500 years ago? Are our starched cotton cheesecloths more abrasive or less? We know that presses were made primarily of wood until the end of the 18th century. How much do our steel rollers and press beds add to wear on a plate?
Even the wiping technique is called into question. I do cold wiping with starched cheesecloth (tarleton), followed by hand wiping with whiting and finished with a bit of paper wiping. Ad Stijnman, whose forthcoming book Engraving and Etching 1400–2000 examines historical intaglio techniques, told me that he has found no documentary evidence of hand wiping with whiting from this period. More importantly, he reminded me that the project of consistently “printing a plate until it wears out,” would have been an alien idea in the 16th century: a good printer would adjust ink, pressure, hydration of paper and wiping technique to get a decent impression even out of a relatively worn plate.
Many of the discoveries from this project are personal: I now know the feel of hammered copper; I have a clear idea of the time and labor required to make a plate; I understand why a lozenge shaped burin does not work well in rolled copper; I have tested myself against Dürer’s engraving of hair and comprehend in a new way the pleasure he must have taken in elaborating this favorite motif, and all of this has changed my approach to my own work.6 But there are objective discoveries as well: the printing process has already revealed markers that distinguish a freshly engraved plate from a matrix that has been polished by wiping and printing. Further printing will provide information about the durability of the plate and chronicle the appearance and disappearance of scratches and stray marks.
This approach will never have the authority of primary research done from original objects as exemplified by Angela’s work on Melencolia I. Nor does it offer the kind of evidence that can be gleaned from documents contemporary to the culture being studied as in Ad Stijnman’s research into printing techniques. Research using studio techniques is speculative and imaginative. But it can bring a historical artist’s work to life in the present in a way that is compelling to audiences at all levels, from the conservators and curators at the Met to the casual museum visitors who have never given any thought to engraving. I hope it enhances our understanding and experience of historical art, and builds a kind of bridge—both cognitive and visceral—across the centuries.
- The project, which initially focused on Melencolia I alone, began as part of a thesis completed in 2008 for an MA, CAS degree in Art Conservation at Buffalo State College, titled “The Albright’s Albrechts: A Study of Micro Plate Wear in 16 Impressions of Dürer’s Melencolia I (1514).”
- Meder, Joseph, Dürer-Katalog: ein Hanbuch über Albrecht Dürer’s Stiche, Radierungen, Holzschnitte, deren Zustände, Ausgaben und Wasserzeichen. Vienna:Verlag Gilhofer & Ranschburg, 1932.
- Campbell, Angela and Dan Kushel. Have Camera, Will Travel: Modifying a Panasonic Lumix Camera for High-Magnification Image Capture and Optimal Portability. http://www.conservation-us.org.
- “The Brilliant Line: Following the Early Modern Engraver, 1480-1650,” Museum of Art, Rhode Island School of Design, Providence RI, 18 September 2009 – 3 January 2010; Block Museum of Art, Northwestern University, Evanston IL, 9 April – 20 June 2010. The drawings were published in Emily Peters, ed. The Brilliant Line: Following the Early Modern Engraver, 1480-1650, Providence: Museum of Art, Rhode Island School of Design, 2009.
- Zoe Wendel, a senior in RISD’s Department of Jewelry and Metalsmithing, made the introduction.
- In the engravings I have made since copying St. Paul the figure scale is much smaller, exactly the size of Dürer’s diminutive but monumental saint. I have not used outlines or crosshatching for many years, but now I am trying out the clear contours and regular fine hatching that entered my work while making the copy. I signed my copy of St. Paul with the monogram I use for my historical interpretations, and I would include it in a catalogue raisonne of my work. It takes its place among the many copies I have made since I first interpreted a 16th-century German woodcut in linoleum when I was 14. I am not quite sure what these copies and interpretations mean in terms of appropriation or contemporary art theory, but I am interested in thinking about it.