David H. Wright

title.none: Williams, ed., Imaging the Early Medieval Bible (Wright)

identifier.other: baj9928.0007.008 00.07.08

identifier.issn: 1096-746X

description.statementofresponsibility: David H. Wright, University of California at Berkeley,

publisher.none: .

date.issued: 2000

identifier.citation: Williams, John, ed. Imaging the Early Medieval Bible. University Park: Penn State University Press, 1999. Pp. 1, 227. $75.00. ISBN: 0-271-01768-6.

type.none: Review

relation.ispartof: The Medieval Review

The Medieval Review 00.07.08

Williams, John, ed. Imaging the Early Medieval Bible. University Park: Penn State University Press, 1999. Pp. 1, 227. $75.00. ISBN: 0-271-01768-6.

Reviewed by:

David H. Wright
University of California at Berkeley

Ignore the ridiculous pseudo-English at the beginning of this title. This book has serious substance, not hot air in imitation of the French. It consists of revised and well illustrated versions of five papers given at a symposium at Pittsburgh in 1994 organized by the editor, John Williams.

The first, by John Lowden of the Courtauld Institute, offers a judicious survey of all surviving illustrated biblical manuscripts dating before the eighth century. He groups them by language (Greek, Syrian, and then Latin) and describes their size and character, giving generally sensible dates and localizations. It is noteworthy that while admitting the lack of firm evidence he favors Constantinople as a reasonable possibility for the three famous purple manuscripts. The case is particularly compelling for the Sinope fragments, an enormous book written entirely in gold letters seven millimeters high. In the Justinianic era it is hard to imagine such a production anywhere but in the capital. It should be noted that on stylistic evidence the closely related Rossano Gospels must be placed a bit later, around the end of Justinian's reign. Also that the extraordinarily luxurious fragment of Canon Tables in the British Library, where gold leaf covered all of each page, matches both in ornament and in the style of painting the busts in the arcades the work of Justinian II and should therefore be dated around 700.

For the Syriac books the famous Rabbula Gospels is firmly dated 586 and is a key monument in the development of Christian iconography. The five other Syriac books, here patiently described, have less to offer and cannot be dated specifically, but it is good to be reminded of them and also of the Armenian Ejmiadzin gospels, and to have some good illustrations.

Coming to the Quedlinburg Itala fragments late in his survey, Lowden gives it only brief treatment and dates it too late, in the time of S. Maria Maggiore (432-40). Its style is clearly earlier than that, essentially the same as the Vatican Vergil, to be dated around 400, plus or minus a decade or so (see my 1993 monograph on the Vergil). It is thus roughly half a century earlier than the oldest known illustrated Greek bible manuscript, the Cotton Genesis (which had 339 framed illustrations placed in the text just after the passage illustrated) and its arrangement clearly reflects uncertainties in undertaking what must have been an entirely new venture. For three of the four surviving illustrations the master of the scriptorium wrote out detailed instructions for the painter; for the other he made rough sketches and wrote a couple of labels, which presumably he supplemented with oral instructions. This is therefore a documented case of what must have been a very large cycle of narrative illustrations invented from scratch. One of the surviving pages has six episodes crowded into four squares; for the four books of Kings there may have been nearly as many episodes illustrated as in the Cotton Genesis, and the Quedlinburg book was comparable in size and length (though squarer in format).

The Quedlinburg illustrations occupy a square frame on a full page located before the text illustrated, but the specific text for the first episode in the first surviving illustration comes on the fourth page after the illustration, and between that illustration and the text of the first episode illustrated are several illustratable episodes, including one, the anointing of Saul, that should certainly have been illustrated. This bad planning seems to be a symptom of the newness of this venture in adding illustrations to a bible and implies that at least in the Latin world the very idea of illustrating the bible was new. The large size, the excellent production standards and the elegant large uncial script clearly point to important patronage. Considering that this is Rome around 400 I have put forward on several occasions (not yet properly published) the hypothesis that this was Innocent I (401-17, known to have taken a special interest in the text of the bible) and that it was one volume in a set, of which I suggest the Genesis was later used as a model for part of the story of Jacob at S. Maria Maggiore, though I am doubtful whether any other mosaics there are based on miniatures. I suggest that papal Rome at the beginning of the fifth century was a great innovator in Bible illustration.

At this point it is worth comparing the situation in the illustration of classical literature. Also from Rome around 400 we have the surviving Vatican Vergil fragments for which I have argued that the Georgics illustrations were invented for that manuscript, and include as the frontispiece to Book 3 a framed page divided into six panels to illustrate the next fifteen verses, an arrangement comparable to the illustrated pages of the Quedlinburg Itala, an arrangement impossible in a papyrus role. I have argued on the other hand that the Aeneid illustrations must have been based on a set of rolls from around the first half of the second century, and also that the illustration to the first Eclogue in the Roman Vergil of the late fifth century must be a relatively accurate copy of an earlier illustration in 'papyrus style' (see my article in Dumbarton Oaks Papers 1996 as well as my 1993 monograph). But I have also argued that the lost Late Antique illustrated Terence (best studied in the Vatican copy) was an entirely original creation made in Rome around 400 (in Medieval Manuscripts of the Latin Classics: Production and Use, ed. Claudine Chavannes-Mazel, 1996).

The second article, by Katrin Kogman-Appel of the Ben Gurion University of the Negev in Beer Sheva, surveys the vexed question of possible Jewish pictorial models and of narrative details in extra-biblical Jewish texts. She distinguishes between Hellenistic Jewish literature (as Philo and Josephus), widely known in Christian circles, and rabbinical writings less likely to have been read in Christian circles. She cites Jewish references to Jewish wall paintings in the third century and to Jewish floor mosaics in the fourth but points out that there is no ancient reference to Jewish manuscript illustration. The Synagogue of Dura-Europos (245) is the only known example of a cycle of Jewish paintings of biblical subjects (as distinct from individual subjects in floor mosaics) and it shows no symptom of direct dependence on manuscript illustration, but it does include a prominent iconographic detail from an extra-biblical rabbinical source (Hiel and a snake in the sacrifice by the prophets of Ba'al).

The search for comparable rabbinical details in Christian art has too often led to special pleading. Do we really need rabbinical learning to lead the painter of the Vienna Genesis to show Potiphar's wife in bed when trying to seduce Joseph? We must reject the infamous proposition that nothing is ever invented, but always derived from an earlier archetype; we must admit that coincidence is always possible, especially when common sense might suggest independent invention. We must also be a little cautious in interpreting minor details that have no labels; is the modestly draped and veiled woman outside Joseph's prison in the Vienna Genesis necessarily Potiphar's wife still trying to seduce him? Identifying such motifs has become a major industry, especially in Israel, but still we have no unquestionable instance of copying from an ancient illustrated Jewish manuscript. Kogman-Appel offers a good guide to these problems and concludes realistically that particularly in Syria and Palestine in the third and fourth centuries there must have been a lot of cultural interchange between learned Jews and Christians. We must not expect to be able to define every such instance precisely.

The third article, by Dorothy Hoogland Verkerk of the University of North Carolina at Chapel Hill, attempts to survey biblical manuscripts produced in Rome in Late Antiquity. She starts with brief consideration of the Quedlinburg Itala (which she dates too late) and the Gospels of St. Augustine (for which a date at the end of the sixth century can be confirmed by comparing the figure style of the mosaic at S. Lorenzo flm). She speculates about the manuscripts that Benedict Biscop obtained in Rome late in the seventh century but does not mention that the only one of these about which we have considerable information is the Codex Grandior of Cassiodorus, which was made at the southern tip of Italy and somehow reached Rome, where Biscop acquired it without knowing its origin. Then she tries unconvincingly to attribute the Ashburnham Pentateuch to Rome. In an earlier article (Art Bulletin 77, 1995, 94-105) she demonstrated that despite the Jewish connections in these illustrations they contain many references to Christian liturgy, but now she attempts to place such features specifically in Rome.

She repeats a mistake of Grabar in comparing the page with Moses reading the covenant with the left niche of Cubicle C in the Via Latina catacomb, where Joshua, carrying on after the death of Moses, with small grisaille scenes above him of Moses receiving the law and of the column of fire as reminders of God's promises, leads the Israelites into the Promised Land, symbolized by the Temple on the mount (as brilliantly explained in William Tronzo's 1986 monograph, which she cites without understanding). There is no significant resemblance between these scenes. She points also to the seven figures in white behind Moses, interpreting them as deacons, and remarks that Rome always had exactly seven deacons. Her other points are no better. That the Ashburnham Pentateuch shows various kinds of resemblance to general Italian practice does not mean it had to be made in Italy, much less in Rome. On the contrary the production quality is very modest and generally suggests a provincial origin. But which province it should be assigned to remains, as E. A. Lowe remarked late in his career, one of the abiding riddles of Latin palaeography. The most popular contenders are Spain and North Africa but no one has found conclusive evidence.

The fourth article turns to northern Europe, where Lawrence Nees of the University of Delaware, writing in a strangely circumambulatory and occasionally egocentric style, first touches on Eusebius's report that Constantine commissioned fifty copies of sacred scripture and then demands a lot of our patience as he lays out a few obvious generalizations about the great Carolingian bibles. Despite the general theme of this book he fails to explain the remarkable originality of some of the pages in the Vivian Bible and he indulges in various scattered remarks that do not significantly advance any argument. Then he turns back in time to the Codex Amiatinus and there he gets into serious trouble. Paul Meyvaert's magisterial article "Bede, Cassiodorus, and the Codex Amiatinus" (Speculum 71, 1996, 827-83) was published after this symposium but in time to be mentioned repeatedly in the finished paper, but Nees has not taken sufficient advantage of its firm conclusions (such as the demonstration that Bede did not know the Institutiones of Cassiodorus) to eliminate discussion of impossible earlier suggestions.

From Meyvaert's work and my own publications nearly forty years ago it is clear that:

1. All the work on Amiatinus was carried out by English monks in Wearmouth-Jarrow, writing on 'Insular vellum'. Nordhagen's claim (1976) to see an Italo-Byzantine painter at work is not worth considering; his expertise is on the wall paintings of S. Maria Antiqua in Rome and he lacked connoisseurship in Insular and other manuscripts. He did not recognize, for example, the important presence of the bright yellow granular pigment orpiment, normal in English work but not in Italian, as pointed out long ago by Nordenfalk.

2. The text was copied in nine separate sections (judging from the relation of text to truncated or extended gatherings). In two instances I have attributed two of these sections to one scribe, leaving a total of seven scribes to account for the whole text, and I have described in detail the different stages of evolution of uncial and Rustic capitals to be observed in the work of each scribe, often including refinements developing in the course of that work (Traditio 17, 1961, 441-56). One of the scribes was a leader in developing the sophisticated hierarchy of classical scripts found in his work and further perfected in the slightly later Utrecht fragments and the Stonyhurst Gospel; other scribes were more or less skilled followers. The work in the first gathering is by a different scribe or scribes, quite advanced in style, and includes some attempt to develop a fluent sloping uncial.

3. The fragments of another of Ceolfrid's pandects now in the British Library are slightly smaller in size and written in slightly more crowded uncials comparable in style to the earlier or more tentative work in Amiatinus. The production of Amiatinus was clearly the culmination of Ceolfrid's book production but given the obvious division of labor in it I do not think it should be supposed it was started more than a couple of years before its unannounced purpose.

4. The dedication page must have been written at the last moment in 716 when Ceolfrid revealed his plan to retire to Rome. E. A. Lowe once suggested to me, and doubtless to others, that because it is quoted in the anonymous life of Ceolfrid but not in Bede's life of Ceolfrid, the dedication may have been composed by Bede, but the style of script on this page cannot be Bede's; it is very formal in character and expertly executed, but not yet as formal as the title page of the Utrecht fragment. A more compressed version of this script was used for the uncial gospel fragments attached to Durham A. II.17. (Meyvaert would like to attribute to Bede some of the less formal writing in the first gathering, note 75 and pp. 869-70).

5. The illustration of the Tabernacle was copied from the Codex Grandior of Cassiodorus but whether it was actually bound at the very beginning of the Codex Amiatinus might be questioned on practical grounds; perhaps for protection and convenience it was placed just in front of Genesis. The three diagrams of arrangements of the bible were also copied with some modifications from the Codex Grandior. The prologue and contents pages and the pentateuch diagram are local work using decorative schemes from Italian books; Meyvaert gives as good a reconstruction of their arrangement as can be devised.

But Meyvaert's treatment of the Ezra miniature is not convincing. With Nees I agree that the notion of the aged senator Cassiodorus having himself portrayed as a scribe at the beginning of his complete bible does not ring true. To require that the Wearmouth-Jarrow artist converted such a portrait into Ezra by adding the breastplate and headdress through antiquarian study asks too much of a painter who made such a mess of the red and green areas of the drapery and of the shadow of the ink bottle. To postulate a Gospel Book in Wearmouth-Jarrow with evangelist portraits to serve as the model for Ezra is also asking too much, for there is good evidence that there was no imported set of Evangelist portraits available in Northumbria at this time: I have argued that for that reason the Lindisfarne Evangelists were invented on the basis of this Ezra (used for both Matthew and Luke), a secular author portrait (used for Mark) and the Christ from a Maiestas (used for John); the Lindisfarne artist must have known about Italian Evangelist portraits in some general way but had no model he could copy directly (see Stil und Uberlieferung in der Kunst des Abendlandes, Akten des 21. Int. Kongr. fur Kunstgeschichte in Bonn 1964, Berlin, 1967, pp. 86-9). We must continue to accept the Amiatinus Ezra as a semi-competent English copy of an illustration in Cassiodorus's Codex Grandior, and while we might imagine that Cassiodorus thought of Ezra as an exemplar for his own activity I do not think we should interpret his illustration as specifically a portrait of himself.

Nees illustrates but does not discuss the Maiestas illustration that comes at the beginning of the New Testament in Amiatinus. Let me add that I have long argued that it must have been based on an Italian manuscript of the middle of the seventh century, for if we look at it sympathetically, ignoring the clumsy draftsmanship, the bulky quality of the figures and the color bands that give a receding space resemble frescoes from the 640s such as the Maccabees in S. Maria Antiqua (Art Bulletin 43, 1961, p. 251, and Bonn Kongr. p. 86).

Near the end of his meandering way Nees looks at the medallion at the top of the diagram of the septuagint division of the bible and speculates that it might have been added to the Cassiodoran diagram at Wearmouth-Jarrow, based on a gold coin of Justinian II showing a bust of Christ in what came to be the Pantokrator type. He does not realize that this coin was struck only from 692 to 695, only in Constantinople, and then was revived in 843 by Theodora as regent for the infant Michael III. But this type of Christ was well known and readily available to Cassiodorus, appearing, for example, at S. Apollinare Nuovo and in the Justinianic icon of Christ on Mt. Sinai. Nees then turns to the John VII fresco above the apse of S. Maria Antiqua where he does not realize that an entirely different Christ type was used, the Syrian type with short beard and hair that appears on the coins from the second reign of Justinian II (705-11). Next he strays into an irrelevant discussion of Anglo-Saxon sceattas (silver coins). This section is ridiculous.

In the last article John Williams of the University of Pittsburgh examines the Bible of 960 in Leon, a famous example of Mozarabic art with 92 narrative scenes from Genesis to Daniel, each placed immediately after the text illustrated. He considers about a dozen of these in detail, repeatedly stressing their unique iconographic features. Two scenes in the Stuttgart Psalter that Mutherich thought might descend from the same prototype for Exodus illustrations as the Bible of 960 are shown to be meaningfully different in details of subject matter. Supposed connections with the sources of the Byzantine Octateuchs are easily dismissed. Other supposed sources are shown not to be persuasive, including two Williams had published previously, and the rarity of many subjects in the Bible of 960 is stressed. On the other hand, there are two cases (Daniel in the Lions' Den and the Sacrifice of Abraham) that are almost universally illustrated, and here there are important similarities between this bible and the seventh- century capitals in San Pedro de la Nave, but Williams is skeptical of Schlunk's citing this relationship as evidence for lost Visigothic manuscript illustration.

Williams reconstructs the procedures used in making this bible written in two columns, how the illustrations often come at the top or bottom of a column of text in order that they can use the extra margin; he makes the case that because the illustrations are very literal and often have very rare subjects they must have been invented by the scribe-artist on the basis of the text he had just written, more or less as the spirit moved him. It is an attractive conclusion, generally convincing in the cases he has discussed, but a lot more systematic study is needed.

It is easy to imagine that this was a very successful symposium in 1994. With the right audience there should have been a lively discussion, especially on the tendency to emphasize originality in the process of manuscript illustration. I have asserted that claim for the Georgics and for Terence in Rome around 400 and my claim was made easy by the complete lack of specific illustrations of those texts in other media. But as a book this volume is a mixed blessing. The first article is a valuable review of the material with thought-provoking insights; the second is a useful review of a vast problem with implicit cautions. The next two should not have been published. The editor's own contribution is enticing; one could wish it were many times as long, fully illustrated, with systematic descriptions to help the unfamiliar reader understand some of the more obscure images. But that must be a separate book; let us hope it comes soon.