Blockbook Editions of the Speculum
For centuries the Speculum humanæ salvationis blockbooks have been a source of speculation and controversy. Until the late nineteenth century they were usually considered to be the work of Laurens Janszoon Coster (1405–84) in Haarlem, where a statue on the market place still honors him with the invention of printing. The editions have been thought, by some, to be the earliest books printed with movable type in a press. But the fact that they were not dated and that the printer and location of the press were not identified have made their position in the history of Western bookmaking hotly contested. It is, however, generally agreed that the Speculum blockbooks originated in the Low Countries.
The Speculum and the Coster Question
The name of Laurens Janszoon Coster first appeared as the inventor of printing in a pedigree of the Coster family which was drawn up in 1559 but supposedly copied from an earlier document. It stated, "Sijn tweede wijff was Louris Ianssoens Costers dochter die deerste print in die werlt brocht Anno 1446" (his second wife was Louris Janssoens Coster's daughter who brought the first print into the world in the year 1446). It appears to imply that it was the daughter who invented printing, but probably the word "die" refers to the father and should be translated as "the one who." The document is preserved by the Haarlem Stadsbibliotheek (fig. V-1).
The humanist poet, engraver, and print-dealer, Dirck Volkertsz Coornhert, writing in 1561, mentions a "very rough" kind of printing being done in Haarlem according to local tradition. He does not name Coster, nor does his contemporary, the Haarlem printer, Jan van Zuren, whose writing on the invention of printing (now lost) was later versified by the learned Petrus Scriverius in an homage to Coster in 1628.
In a treatise by Hadrianus Junius entitled Batavia (an old name for Holland), written in 1568 and published at Leiden in 1588, Coster is credited with the invention of movable characters and with printing a Speculum in Haarlem about 1440. The Hadrianus account of Coster's invention stated that on Christmas Eve, when everyone was at church, an apprentice of Coster's stole his type and equipment and took it to Mainz, where he printed a Doctrinale with the same characters. As for the invention of the method, Hadrianus relates that Coster first cut single letters in the bark of a beech tree (perhaps in the immemorial custom of cutting initials on trees), but had the inspiration to lift them out and print words from their reverse sides. Through some method of casting he was supposed to have created metal types with which he printed whole pages on paper or vellum. Supporters of the story have cited the statement of the veteran printer Ulrich Zell on the invention of printing in 1440 to the author of the Cologne Chronicle (Johann Koelhoff II, Cologne, 1499), which appears in that volume as "die eyrste vurbyldung vonden in Hollant vyss den Donaten" (the first prefigurations were found in Holland in the Donatuses). The study of the earliest Dutch printing shows that there were, indeed, Donatuses printed, but there is no clear proof to connect them with Haarlem or Coster.
Dating, Localization, and Attribution
Many writers, published as long ago as Gerard Meerman in 1765 and as recently as Wytze and Lotte Hellinga in 1972 and 1973, have explored the problems of the printer, place, date, and sequence of the Speculum editions. In 1863, William Young Ottley summarized the major earlier studies and added his own extensive research, but the questions continued to remain unresolved. In 1871 Henry Bradshaw of the Cambridge University Library proposed a date by citing the only one in a first edition of the Speculum blockbook, 1471, written in red in a fifteenth-century hand in a Munich copy (Universitätsbibliothek München, Cim. 52, Xyl. 10). Since then the first edition is usually listed as "not after 1471." He also located the printer in Utrecht, based on the use there of two of the blocks by Jan Veldener in his edition of the Epistelen en evangelien in 1481. How and when Veldener obtained the blocks is not known, but their use
does not locate the printing of the text in Utrecht, and, indeed, the blocks may never have belonged to the text printer. In the British Museum Catalogue of Books Printed in the XV Century , Vol. IX (BMC IX), the products of the press are assigned to the "Printer of the Text of the Speculum." Perhaps the anonymous printer wished to follow the example of the author of the manuscript, and out of humility, to remain unknown. In any case, this identification has come to be used to designate the producer of the early Netherlandish printed works formerly referred to as "Costeriana."
The same types as those in the Speculum were used extensively, as well as related type faces, in Doctrinals (fig. V-2), Donatuses, and Distichs, of which fragments continue to be found (fig. V-3), but none, as yet, with a date, printer's name, or location. Editions in these types (with certain other features), which are generally accepted as the most ancient to be printed in the Low Countries, are called Netherlandish prototypography. The only date on any other example is an owner's inscription of 1472 on the Darmstadt copy of De salute corporis . This book is set in a Dutch blackletter, named for the author as the Saliceto type. It is similar to the Speculum type, but larger. The printer of the Speculum had two other types, the Valla, for printing Laurentius Valla's translation of Aesop, and the larger Pontanus type, used in that writer's Singularia in causis criminalibus . These types are grouped together because of their use in similar texts and their likeness of forms.
The only documentary evidence of a press in the Low Countries near the middle of the fifteenth century is in the inventory made at the death of a nun of the Bethany convent at Malines (Mechelen), Jacoba van Loos-Hensberghe, in 1465. This states that she had inherited from her brother "unum instrumentum ad imprimendas scripturas et ymagines " (an instrument to print writing and pictures). The inventory also included novem printe lignee ad imprimendas ymagines cum quatuordecim aliis lapideis printis " (nine wood blocks for printing images, with fourteen other stone blocks). Could Jacoba's late brother have been our mysterious, anonymous printer? No. He was the Bishop of Liége.
Several scholars have made studies of the possibility of a printing workshop in Malines in the 1460's. Three prints, and one leaf combining xylographic and typographic printing, which have been attributed to the Convent of Bethany, have been preserved. The leaf is headed by a woodcut carrying the title Ecce panis angelorum and it is in the Brussels Royal Library. However, based on a study of the characters, it has been concluded that the original could not have been printed before 1530. It is generally accepted that the press which printed the text of the Speculum editions and the schoolbooks in Speculum types was located in the northern Low Countries, but it is not known in what city it operated.
The Speculum Editions Compared
The types used in the editions of the Speculum have been described in BMC IX and in the work of the Dutch scholars Wytze and Lotte Hellinga. The measurement of the characters is based on millimeters required for twenty lines of text from the base of the first to the base of the twenty-first line. The letter G, in the following descriptions, refers to Gothic, a blackletter form of type.
Four editions were produced, two in Latin with verses in rhymed couplets, and two in Dutch prose translations. The order of their production was debated for several centuries, but the accepted sequence today is as follows. The dates given for the editions are those deduced by Allan H. Stevenson from his study of the watermarks of the various papers used.
I. First Latin edition (c.1468), with blocks in good state and Hellinga Type 1: 110 G (figs. V-4, 6, 7, 8).
II. First Dutch edition (c.1471), with block borders occasionally broken, in Hellinga Type 1:110 G, but with two pages printed in another type, Hellinga Type 2:103 G (figs. V-5 and 10). The Latin captions in the woodcuts are translated into Dutch at the head of each column.
III. Second Latin edition (c.1474), with the blocks showing further deterioration, in Hellinga Type 1, still in good state. Twenty pages of this edition have text in xylographic imitation of the first Latin edition (fig. V-9).
IV. Second Dutch edition (c.1479), with the blocks even more damaged, and the text in Hellinga Type 1, but cast on a smaller body. The characters are badly worn, poorly printed in a sooty black ink, and unevenly aligned (fig. V-11). This led some writers to think that this was the first edition.
Unlike the complete manuscripts, the blockbook editions do not contain a Prologue, and the Prohemium precedes the illustrated section. The Prohemium occupies four or six leaves (the latter including a blank leaf), according to the edition. The rest of the book is made up in three gatherings of fourteen leaves and one of sixteen.
Only twenty-nine of the forty-five manuscript chapters are printed in the blockbooks although the blocks were made for many more. The text follows the sequence of the manuscripts for the first twenty-four chapters. Then only five are added, possibly to sell the shorter editions more rapidly at a lower price, or perhaps because of the limits of available paper. Some idiosyncracies in each of the four editions shed light on early printing, and these are examined below.
Edition I : In this edition the first five pages are devoted to a Prohemium or summary, in rhymed prose, of the contents of each chapter (fig. V-6). This ends with a statement about the usefulness of the Prohemium for poor preachers who cannot afford the entire book. Presumably the Prohemium was distributed separately (it is a single gathering of the book), but no examples of it have been preserved as an entity. At the end there are two lines about the Second Joy of the Virgin Mary, the Visitation from her cousin Elisabeth, the queen of Assyria, which, since it occurs in Chapter XLV of the manuscripts, does not appear anywhere in the blockbook editions (fig. V-7). Perhaps the manuscript from which the printer worked, which would have had all forty-five chapters, accidentally lacked these lines, and the scribe wrote the necessary insertion at the bottom of the page. The compositor may not have noticed the place which had been marked for insertion in the manuscript and simply added the lines to the last column. This error appears also in the second Latin edition but was rectified in the Dutch prose translations.
Another curious printer's error appears in the first Latin edition as compared to Latin manuscript Clm 146 in Munich. It is the addition in Chapter II, at the head of the third column, of a line which should be line 51 of Chapter VI: Maria autem viro in desponsatione jungeretur . The correct line 51, Chapter II, is: In omni enim re semper debitus modus est tenendus , which appears as line 52. How could such a mistake happen unless the printer were tipsy, or, starting work before dawn on a winter morning by the light of a single flickering candle, he accidentally opened his manuscript to the wrong page, and was too drunk with sleep to notice his error? He has also omitted the word autem and made arbitrary contractions of the Latin words. Furthermore, in order to make up for the extra line and still retain the standard twenty-five line column, he took the liberty of condensing two lines into one, using most of line 61 with the last two words of line 62, the rest of which is omitted, thereby breaking the rhyme scheme. In Chapter VI, where line 51 is missing, in order to fill the twenty-five-line column, the typesetter ended it with the line which should begin the next column. Then to bring that one up to proper length, he introduced an unrhymed line, which does not appear in the manuscript at all, after the first couplet. In fact the width of the column is too narrow for the type, which causes excessive contractions and lines broken over to the ends of succeeding lines. A "poor preacher" would have had trouble deciphering the text but might have relied more on the pictures to inspire his sermons.
The woodblock titles cut beneath the pictures are transposed in Chapter XXII c and d. The caption for the Spies Carrying the Grapes appears beneath the scene of the Killing of the Heir to the Vineyard, and vice versa . This seems to show that the letter-cutter was not the cutter of the pictures.
Edition II : The second edition was produced by rubbing the same blocks, but with the printed text translated into Dutch prose. This edition contains a different oddity. The text is printed in the same characters as the preceding Latin edition, Hellinga Type 1 : 110 G, except for the two conjugate leaves, Chapter XXIII columns a and b (fig. V-11), and Chapter XXVIII columns c and d. This leaf is printed in a smaller, battered type known as Hellinga Type 2 : 103 G. Did Type 1 become unavailable, or tied up in another book, thus forcing the printer to use a worn-out face to complete the edition? Or did he accidentally start setting from a wrong case, and continue when, or if, he discovered his mistake, because the types were similar?
Edition III : This is a reprint of the Latin text of the first edition, but the type has been re-set. It is certainly the most baffling of all four. It contains twenty pages of text cut in wood from tracings of the printed pages of the earlier edition (fig. V-9). These are printed by rubbing on the back of the paper, with the text blocks coated by the same watery ink and rubbed at the same time as the illustration blocks. To us, the most logical explanation for the twenty xylographic text pages is that the making and rubbing of the illustration blocks was done first, and in another location than the printing of the text.
Our hypothesis is that the blocks were rubbed onto the blank sheets, which were then delivered to the text printer (who may never have seen the blocks themselves). When the printing of the text was completed, they were returned to the original workshop for gathering, binding, and distribution. In the case of the second Latin edition twenty of the sheets may have been damaged or lost in transit between the text printer and the house where the books were put together. Instead of preparing a new set of woodcut sheets of illustrations to be sent to the text printer to complete the edition, the missing texts were cut in wood and both the blocks and the texts were printed together by rubbing, an economical solution in time and expense. Since no black type-printing ink was available there, both woodcuts and text blocks were rubbed in the watery yellow-brown ink which had been used for the pictures. There must have been a copy of the Latin first edition on hand, for the text was traced on a transparent sheet, or the actual printed sheet was made transparent with oil, so that it could be flopped and fastened to an uncut block for cutting the characters in reverse. The cutting corresponds exactly in size and spacing to the printed pages.
Edition IV : The printer must still have had the matrices for the missing type for the twenty pages, for he evidently recast the characters on a different body size, and used the font heavily, possibly for the schoolbooks of which so many fragments have been found. It appears, much worn, in the second Dutch edition. The smaller body of Type 1 (fig. V-12), which we examined in the Huntington Library in California (formerly the Pembroke copy), measured 106 mm. instead of 110 mm. for twenty lines. Could it simply be exposure to the dry Southern California climate which caused the paper to shrink? Alas, such an airy explanation is contradicted by the fact that most pages of one of the copies, Haarlem Inv. II, No. 14, which we measured in 1979, required only 104 mm. for twenty lines. This edition is usually referred to as the "unmixed" Dutch issue, but in Haarlem, Inv. II, No. 15, there are actually four pages in 110 mm. type. Pages sixteen through nineteen are missing. We assume that pages twenty through twentythree were taken from the first Dutch edition and bound into this copy, but evidently pages sixteen through nineteen were not available. The other Haarlem copy, Inv. II, No. 14, had also been taken apart some time in the past, and the pages were mounted on larger sheets.
There is another oddity in this fourth and last edition of the Speculum blockbook. In Chapter XIX, column two, the biblical reference line "Genesis IX capitel" is printed upside-down in all known copies (fig. V-12). This might indicate that the characters within the line were attached to the text by some method too difficult to correct. The question will be dealt with further in the section on the printing, below.
Type Printing in the Fifteenth Century
Whether early interchangeable type characters in Europe were cast as flat metal plates or on more easily held shanks of about the 2.5 cm. height of modern types has long been a source of speculation. The method of holding a page of type together is also uncertain. There is a possibility that letters were composed and held in place in a metal tray or galley coated with beeswax and placed on the bed of the press. Heating the back of the galley would facilitate the removal or replacement of the characters.
A similar method appears to have been used in Korea from the eighth century on. Ancient Korean wood types, about 1.25 cm. high, survive They taper slightly from top to bottom, which would have permitted wax to push up between the letters when they were set in a wax-coated tray. The surfaces of the letters could then have been levelled by tapping with a block and mallet. This would have held them securely and compensated for differences in their heights. Such type forms still exist in the National Library in Seoul. But as yet no evidence of the beeswax "lock-up" method of the fifteenth century has appeared in Europe. In Korea and China lines of type were known to be held together with thread, string, wire, or bamboo strips through holes in the bodies of the characters. In 1868 some metal type characters which had holes in the shanks were fished out of the Saône near Lyon. For a time it was believed this showed that the same tying-up method was used in Europe, but it was subsequently demonstrated that the holes were due to faulty casting.
An ancient practice in the early printing houses was to wind twine around the sides of the whole page, as is done today for quick proofing and for storage. This method, if it was used for prototypography, would explain the sometimes irregular alignment of characters. It might also account for the upside-down line at the foot of a column in the second Dutch edition of the Speculum . The printer could have decided it was too much trouble to untie the form when, or if, the error was discovered. If the text was printed from cast metal plates, the line would have had to be sawed off, turned around, and remounted, no small task. However there is no evidence preserved of such plate casting.
We believe that individual characters on rectangular shanks, cast in adjustable molds, were used for the Speculum . As for the invention, the anonymous printer could have been experimenting concurrently with Gutenberg. History is full of independent discoveries in different
places simultaneously, and movable type had reached its time. The letters could have been fixed in a frame by the simple pressure principle of opposing wedges. Straight strips of wood and wider blocks placed between the wedges and the form would have filled out the frame. We know such strips were used in printing the Speculum , between the text columns, because on several pages they accidentally became inked and printed.
In the Jost Amman woodcut of an early press (fig. V-13) the major parts can be clearly seen. At the right, the surface of the type, on the bed of the press, is being coated by means of ink balls. In the center is the tympan, a frame stretched with paper or parchment. The printed sheet was correctly stationed by being pierced through points in the tympan. This is hinged to the press bed on one side and on the other to a frame called the frisket. Holes were to be cut into the sheet stretched across it, to expose the printing areas and mask out the blank spaces. The frisket was folded over the tympan and held the paper in place. Then both were folded onto the type form and all were pushed halfway beneath the press for printing the first page, and all the way for the conjugate page. One can see the bar at right which was pulled by hand and turned the screw to press down the platen and print the page. Sometimes the frisket was miscut and the characters used to fill out the type lines or column depths were also printed.
The papers used for the Speculum blockbooks were small folio sheets about 30 × 44 cm., which barely permitted the rubbing of two adjacent woodcuts, 19.5 × 10.5 cm., across the top of the sheet. The same papers are found also in dated civic records of northern European cities, and they have been used as evidence for dating the printing of the various editions. They were generally of good quality and served well for printing the blocks by friction, though not well enough to prevent the water-based ink from coming through, and therefore the paper was printed on only one side.
Paper was made by dipping a rectangular wire screen mold, with its deckle frame, into a vat of rag pulp, shaking the mold to mesh the fibers, and depositing it on a pile with interleaved felts to absorb the water. Then the pile was squeezed in a screw press and the sheets were separated and dried.
Two molds with the same watermark design were used alternately by the workers, to keep the process continuous. They had a "laid" formation, with closely spaced wires (about six per centimeter) in the long direction and cross-wires, called "chain-lines," woven at wide intervals (about 4 cm. apart). The design of the watermark, shaped in wire, was sewn to this screen by loops of wire thread which often show as bright spots in the outline when a light is placed behind the sheet. The design can be traced (though this method is often unreliable), or it can be photographed at a one-to-one ratio. A better method is the beta-radiograph which uses an isotope plate with an X-ray film. This gives an accurate image of the variations in the thickness of the paper caused by the wires, chain-lines, watermarks, and sewing dots, but not of the printing on the paper. In 1970 Thomas Gravell developed a method which was quick, inexpensive and equally accurate. It uses Dupont Dylux 503 photo-sensitive paper, and exposure through Diazo fluorescent lamps. Electron radiography is also being used in short exposure times to make very sharp images. The precise images produced by these techniques can show the shape and position of the watermark and sewing dots which are caused by the repeated shaking of the mold. This may reveal the sequence in the production of the sheets and provide evidence for dating, when the mark is identical with one in a dated sheet.
We have no way of knowing how many copies of the first Latin edition, or any of the others, were produced. Considering its scarcity today, the number may have been small. Only nine complete copies of the first edition have survived, of which we were able to examine six (see note 13). Probably the house where the woodblocks for the proposed edition were made would have bought just enough paper for current use and sale. The printing of the text must have involved another expense. Only when the pages had been returned from the printer, gathered, glued back to back, and probably bound, could the books be sold and the investment made for more paper and printing another edition.
Edition I : Each edition of the Speculum had a different basic "running" paper in the illustrated part of the book, usually beginning with the first chapter. When that was exhausted similar sheets were purchased in small lots from a paper merchant, or sheets of the same size and quality which were left over from other books were used to complete the edition. The excellently preserved copy of this edition in the New York Public Library has five gatherings: one of three sheets, three of seven sheets, and one of eight sheets, or thirty-two sheets in all. It is collated a:6, b-d:14, e:16, the numbers referring to leaves. The portion with woodblocks, in gatherings two through four (b-d), is printed on a paper with an Anchor and Cross watermark. But the fourth gathering contains a single sheet with a Unicorn. The fifth gathering starts with an Anchor sheet, but the remaining seven inside sheets have Unicorns. Finally, the first gathering, containing the Prohemium, which was printed last (when the number of chapters was known), begins with a Unicorn, but the two interior sheets have Bull's Heads.
It would appear from the evidence of the watermarks that enough paper was bought with the Anchor and Cross device for the first twenty-one chapters, or three gatherings of seven sheets, or fourteen pages. After the rubbing was completed, only eight more chapters were added, instead of the usual twenty-four more found in the manuscripts. Perhaps the paper merchant had in stock just Unicorn watermarked sheets. There was enough to print the seven interior sheets of the last signature. The Anchor and Cross sheet was used as the first wrap-around sheet for the fifth gathering. The Unicorn paper was also used to print the first sheet of the Prohemium and another few sheets were used to make up a shortage in the second sheet of the fourth gathering. (In the Haarlem copy this sheet has the normal "running" paper Anchor mark.) But two more sheets were still needed to complete the Prohemium, and for this Bull's Head stock was found. The limited paper supplies at the time normally precluded having the same watermark throughout an edition, and extra quantities were purchased as needed.
Edition II : The next printing was the first Dutch edition, of which we examined two of the four surviving copies (p. 118, n. 13). It had as its "running" stock a different Unicorn; this time the beast is found vertically on the page. In the British Library copy, the text of the first gathering, a two-sheet, four-page Prohemium, is printed on the Unicorn stock, as are the next three gatherings of seven sheets each. One sheet of this copy, the fourth in the second gathering, is on paper with a large Gothic P. The fifth gathering begins with a Unicorn sheet, but then comes a mixture of three Bull's Heads and four more Gothic P's. One Bull's Head is different from the others. This is the sheet used for the printing of the text with Hellinga Type 2:103 G. It is in a battered condition, apparently because neither the previous Bull's Head stock nor the original type, Hellinga 1, were available.
Edition III : An analysis of the watermarks of the papers of the second Latin edition shows a great variety in both the pages combining woodcuts with type and those which are entirely xylographic (see note 13). The basic "running" stock has a large Y with a tail that curls into a heart. Briquet found it only in this edition of the Speculum , but it also exists in the account books for 1475 in the Gemeente Archief in Haarlem. This mark was apparently made in honor of Yolande II when she became Duchess of Lorraine in 1473. Other marks are scattered with no apparent pattern through different copies. There is another Y with a trefoil tail, probably also intended to celebrate the Duchess. It is also found in the third edition of the Apocalypse blockbook in Copenhagen.
The Anchor and Cross watermark makes a connection with the "Printer of the Text of the Speculum." The twin marks appear several times in his printing of the Pontanus Singularia (Hellinga Type 4:142 G). They are also found in a Biblia pauperum in the British Library and a third edition of the Apocalypse in the Musée Condé at Chantilly. The Y-trefoil mark appears in his Saliceto Tractatus de salute corporis (Hellinga Type 5:123 G), as well as the Pontanus.
Edition IV : We examined three copies of the second Dutch edition with the type re-set but very worn (see note 13). The "running" paper is a Bull's Head with a shield of Metz and a star above it. Several other marks make their appearance haphazardly. There is a Gothic monogram for Maria (ma[*] ) in a circle with a shield of Metz below. The same mark appears in Veldener's printing at Utrecht of the Fasciculus temporum dated 1480. There are also a Gothic P and two Hands of Blessing in the Spieghel edition.
Of the above watermarks which appear in Briquet's Les Filigranes , the dates given are earlier than those assigned by Allan Stevenson from his studies of the watermarks of the blockbook editions. Unfortunately it is not always possible to tell if Briquet's tracings are identical with our beta-radiographs. Precise dating by watermark is still questionable.
The same illustration blocks are used throughout the four editions. Their shape and dimensions clearly indicate that they were intended to be part of the pages of a book, and each is captioned in Latin cut into the block. The drawing, cutting, and rubbing of the blocks could have been done much earlier than the printing of the text and could, in sequence, have been separated by periods of time.
As determined by differences in the vertical measurement of the blocks on conjugate pages, a separate double-image block was cut for each page. The smooth framing arch and round column bases of the first two pictures of each chapter were probably designed to indicate that they were to be printed on the versos, or left pages, and the broken arch and plane-based columns indicated the rectos, or last half of the chapter (see figs. V-11 and V-12).
The few banderoles which were cut away to the borders, Chapters XXV b (p. 190) and XXIX d (p. 199) must have been left blank to be inscribed by hand. Would this indicate that the blocks were designed for a proposed chiro-xylographic book? But the letter-cutter has incised the lettering in the earlier block of the Annunciation in blockbook Chapter VII a (p. 154). In the copy of the Latin first edition in the New York Public Library (formerly the Inglis copy), and the one in the National Library in Florence, the banderoles of the last woodcut, Chapter XXIX d, for "Mene, Mene, tekel upharsin," are not cleared out and print as solid bands. This must have been corrected, after a few rubbings, by clearing away to the borders as is shown by other copies of this edition. It is most unlikely that a chiro-xylographic book was planned, for the labor of writing the extensive text, as well as lettering the banderoles, would have been excessive for an edition of even moderate quantity.
It is generally agreed that there were two artists or woodcutters at work on the series. The first designed the pictures for Chapters I through XXIV, the next manuscript chapter is omitted, and in Chapter XXV of the blockbooks, a new artist is seen, or possibly only a new cutter. The architectural framing of the scenes and the solidly centralized or diagonally balanced compositions are very like those of the earlier chapters, but the form of the trees is rounded and the bodies are stockier, with broader heads. The hatchings for shading are frequently diagonal rather than horizontal. Finally, the figures are often larger within the frame.
Drawings were made, and blocks cut, for more of the manuscript chapters (if not all forty-five) than were used in printing the twenty-nine chapters of the blockbooks. Jan Veldener sawed in half all of the woodcuts (58 for 116 pictures) which appear in the books and used them in his Culemborg edition of the Speculum in 1483, together with eleven more by the artist/cutter of the later blocks (see our Chapter VII).
The first artist of the woodcuts either was a known Netherlandish master of the period or worked in his atelier. Mention has been made in Chapter I of the magnificent Book of Hours
of Catherine of Cleves, which most art historians agree was completed about 1440. The manuscript contains several miniatures which in their style and subjects show that the illuminator was aware of the woodcuts of the Biblia pauperum and the Speculum humanæ salvationis or of lost models which may have been used for both the blockbooks and the miniatures in a closely knit community of book producers. There is a version of the goat nibbling the grapes in the Speculum woodcut of the Shame of Noah (Chapter XIX c) which in size and drawing is very close to the illumination in the Cleves manuscript.
The possibility must also be accepted that the artist of the blockbooks was influenced by the Cleves miniatures rather than the reverse. Since the workshop where the Hours was made is known to have continued production through the middle of the century, it may be that an artist from the circle of the Cleves Master was actually involved in designing the woodblocks of the Speculum. We may hope that further explorations by art historians and bibliographers will reveal more about the connection between the contemporary manuscript illumination and the Speculum blockbooks.
A relationship has long been assumed between the Altarpiece of the Blessed Sacrament painted by Dirck Bouts (c.1415–1475) for the church of St. Pierre in Louvain and the blocks for Chapter XVI of the Speculum . In 1464 a contract was executed by which the Confraternity of the Holy Sacrament of Louvain, a group of wealthy burghers organized for pious and charitable purposes, commissioned Dirck Bouts to paint a triptych for the larger of their two chapels in the choir of St. Pierre. It was to be dedicated to the Eucharistic rite, which had enormous popularity at the time, and the Brothers chose two theologians to inform Bouts on the subjects and arrangement of the panels. They based the iconography of the triptych on the Speculum humanæ salvationis , which had been widely accepted as an authoritative work for more than a century.
Bouts' triptych depends on the pictures of the blockbook not only in the selection of the subjects but also in the general compositions of the five scenes. They have been enriched by many details. Portraits of the theologians and possibly the artist himself are included in the central panel of the Last Supper. In the Eating of the Paschal Lamb, on the table are the type of glasses made in the Meuse and Brabant regions. The painting of the Gathering of the Manna draws its concept of the Heavenly Bread from the biblical description: the manna is "small as the hoar frost on the ground" (Exodus XVI: 14). The pitchers used for gathering it are of the same shape as those in the Speculum blocks and are probably typical of that period and location. If the Bouts triptych did depend on the blocks (and we know of no manuscript miniatures with these same compositions), the dating of the first edition at 1468 on the evidence of watermarks is thus brought into question, for the blockbook must have existed before the 1464 contract.
The triptych of the Blessed Sacrament is thought to be the first panel painting of this subject to be done in the Netherlands and its focus on the act of Communion, in the central panel, was unusual. Bouts appears to have worked with Roger van der Weyden on the angels for the vault of one of the chapels of the church of St. Pierre before undertaking the commission for the altarpiece. Both Bouts and Van der Weyden are known to have had strong relationships to the Groenendael community of the Devotio moderna .
We have discussed in Chapter IV the production of blockbooks, specifically the Spirituale pomerium and the Exercitium super Pater Noster , at Groenendael, or possibly at Sept-Fontaines, or both. We have outlined the great importance of the many houses associated with the Devotio moderna in fifteenth-century book production in the Low Countries. We know that they contained woodcutters, illuminators, scribes, miniaturists, and binders. They were also, we assume, in somewhat the same position as a publishing house, i.e., they chose the texts, selected the artists or ateliers for the illustrations, commissioned the text printing, and distributed the editions. The production of two of the editions in Dutch, and the use for which they were intended, strongly suggests that the blockbook editions of the Speculum can be linked with the houses of the Devotio moderna . It also seems logical that the woodblocks remained in the possession of one of the communities (probably Groenendael or Brussels) where Veldener could have seen them when he set up his press nearby in Louvain, in 1475, or before he left for Utrecht in 1478 (see our Chapter VII).
While the name of the printer, the place, and the dates of the printing of the Speculum blocks or text cannot, as yet, be determined, the continued research of art historians, bibliographers, and typographers may combine to solve these mysteries.