The quincunx: its origins and development in the art of Western Europe in the second half of the 8th – the beginning of the 11th centuries

Размещено на http://www.allbest.ru/

Федеральное государственное автономное образовательное учреждение высшего образования

«Национальный исследовательский университет

«Высшая школа экономики»

Выпускная квалификационная работа

The Quincunx: Its Origins and Development in the Art of Western Europe in the second half of the 8th - the beginning of the 11th centuries

О.С. Воскобойников

Mосква, 2020

Table of Contents

Introduction

Chapter 1. The Quincunx Paradigm: Introduction to the Subject

1.1 Quincunx as a word and as a term

1.2 The Roman origins

1.3 Quincunx in the context of multi-medallion compositions

Chapter 2. Origins of the Quincunx in Christian Art

2.1 The first examples of the quincunx scheme in Christian art

2.2 The case of the 8th century Maiestas Domini

Chapter 3. Quincunx in Religious and Scientific Iconography in the 9th-11th centuries

3.1 Quincunx as a scientific diagram

3.3 From Carolingian to Ottonian tradition: the sophistication of the quincunx

3.4 Quincunx in the 12th-15th centuries: an overview

Conclusion

Bibliography

Introduction

The quincunx is a term that describes a centered, symmetrical five medallion scheme. It can be applied to a pattern, a scientific diagram, or a composition of an artwork. In the Middle Ages, this five-medallion scheme was applied to a variety of religious subjects, but most prominently to the iconography of Maiestas Domini. The issue is where the scheme came from, what its purpose and meaning is, and how the implementation of the quincunx composition influences the subject that it comprises?

The aim of this work is ambiguous: we are striving to both identify the origins of the quincunx scheme, and to trace the key points of its development in the early Christian period. In order to fulfill this challenging purpose, we are undertaking three main research objectives. The first is to determine the point in time when the quincunx emerged in the realm of ornaments and figurative imagery, and to define its basic qualities as a pattern and as a composition. The second is to track down the process of its appropriation by Christian subjects, and to ascertain the mutual transformations that occurred within this new scheme-subject relationship. The third is to detect when and why the quincunx scheme peaked as the composition for Christian iconography, and how it functioned in the early medieval period, particularly during the Carolingian and Ottonian eras.

These research objectives determined the structure of our work which consists of three chapters. The first one represents the introduction to the subject where we will be treating the most fundamental questions regarding the quincunx scheme and its origins. Among them is the genesis of the quincunx as a word and as a term. This part is substantial in terms of understanding how the quincunx as a visual structure was perceived in different time periods. Furthermore, in this chapter, we will look at the first examples of the quincunx scheme in the pagan Roman context and will try to understand in what circumstances the quincunx appeared and consequently disseminated in the late antique, early Christian, and early medieval realms. In the second chapter, we are dealing with the quincunx scheme as it is applied to the Christian subjects. There, we are analyzing the process of structural changes that underwent the quincunx composition. Our primary concern will become the iconography of Maiestas Domini and its first example in the quincunx form, namely the Gundohinus Gospels. By examining the reasons of the quincunx composition being applied to this particular subject, we aim to uncover the main characteristics of this scheme. In the third and final chapter, we will see how the quincunx scheme, during the Carolingian and Ottonian periods, will evolve to embrace both scientific and religious notions.

Therefore, the subject of our work is the origins and development of the quincunx composition during the early medieval period. Particularly, we are interested in the timespan from the middle of the 8th century when the quincunx composition was applied to the iconography of Maiestas Domini for the first time, to the beginning of the 11th century when the quincunx scheme significantly changed in terms of both the visual appearance and its meanings. The objects of our research are all the examples of the quincunx schemes, starting from the 1st century AD onward, in all possible media.

During the initial stages of our research, we looked through and analyzed literally tens of thousands of images in order to find the quincunx patterns and compositions in the late antique, early Christian, and early medieval art. Eventually, we ended up with approximately four hundred artworks of which only about one hundred will be used in this work. What are the principles of such a selection? The objects, i.e. our primary sources, might be divided into three main categories according to our research purposes and the structure of this work. In the first chapter, we will concentrate on the quincunx schemes that represent ornaments and patterns which primarily occur on the mosaic pavements, coins, jewelry media, and in some of the book illuminations, specifically in Anglo-Saxon context, a good example of which may be the Lindisfarne Gospels. In the second chapter, we will be using the two kinds of sources. The one group is represented by works selected on the basis of their `transitional' nature, meaning that they are a part of the context in which the pagan culture encountered the Christian tradition. And thus, we will be able to witness the transformation from one framework to another. Among this type of works are catacomb paintings and some of the textile pieces. The other group is represented by a wide range of objects that somehow related to the center artwork we are analyzing in this chapter, namely the Gundohinus Gospels. There, we are tracing the origins of its unusual composition with respect to all the media that might have influenced this illumination, from stone reliefs to vault decoration to the pearled borders of the byzantine coins. And finally, in the third chapter, the choice of primary sources is resulted from our main objective to specify the nature of relations between scientific and religious iconography within the quincunx composition. Therefore, the selection of the works implies first and foremost the medium of manuscript illuminations where such a liaison between scientific diagrams and religious composition originally occurred. Basically, due to a relatively modest amount of preserved examples, we will be engaging most of what we were able to find in our analysis.

In our methodology we are drawing on several sources. To study a composition, one needs to use multifaceted methods. Firstly, because the quincunx has a formal aspect, and especially because it might represent a pattern, we need to study it formally, like a geometric scheme. However, a medieval composition cannot be examined only by means of its visual structure: the art of the period is too closely connected to the religious notions and symbolism. Thus, it needs to be placed in the context of its subjects. Therefore, we are going to implement an iconographic method formulated by Male Mвle Й. L' art religieux du XIIe siecle en France: Etude sur les origines de l'iconographie du Moyen age. Paris: A Colin, 1922; Mвle Й. L'art religieux du XIIIe siecle en France: Etude sur l'iconographie du Moyen Age et sur ses sources d'inspiration.Paris: Leroux, 1898. and, most lately, Baschet Baschet J. L'iconographie mйdiйvale. Paris: Gallimar, 2008., in order to compare all the themes that were inhabiting the quincunx scheme in different time periods. This, hopefully, will lead us to an understanding of the compounds of meanings represented by such a form. Furthermore, since our aim not only to grasp the meaning of the quincunx composition but also to track its deeper roots, we are going to use a method of genealogy worded by K. Weitzmann and H. Kessler The Illustrations in the Manuscripts of the Septuagint / Ed. by K. Weitzmann, H. Kessler. - Princeton: Princeton University Press, 1986.. And finally, because our third chapter is intended to demonstrate the correlation between scientific and religious notions within the quincunx paradigm, we will be drawing on the methodology that was developed almost twenty years ago by Bianca Kьhnel Kьhnel B. The End of Time in the Order of Things: Science and Eschatology in Early Medieval Art. Regensburg: Schnell and Steiner, 2003.. She compared similar structures and their meanings and, studying the context in which such a similarity occurred, she was able to make conclusions regarding the complex relations between art and thought of the period. Her methodology is both highly valuable and ambitious, and we will try to follow her footsteps at least in some aspects.

Review of secondary sources

The study of the medieval compositions, and particularly the study of relations between the structures in religious iconography, ornaments, and scientific diagrams, to which we are aiming to devote a core of our research, is not the most extensive phenomenon in the art historical or medievalism realms so far. Particularly, the study of the origins of such a schematic and unusual composition as the quincunx started only seventeen years ago after the publication of the aforementioned Bianca Kьhnel in 2003. What was the context in which this monograph was able to emerge?

Before Kьhnel, there were approximately seventy years of study of the medieval compositions of various forms. The most attractive to art historians turned out to be the concentric circular scheme to which one of the first publications on the subject was devoted. Initially, the problem of the origins of the concentric composition was developed by J. Baltrusaitis Baltruљaitis J. L'Image du monde celeste du IX au XII siиcle // Gazette des beaux arts. Vol. 20. 1938. P. 137-148. , and he was the first to suggest the correlation between this form and the celestial and cosmological notions. He analyzed a group of the manuscripts of the 11th and 12th centuries with circular compositions and connected the representations of heavenly visions and theological formulae such as Hortus Deliciarum with the actual illustrations to the scientific knowledge of the period. Furthemore, he was the first to trace these circular schemes far back to Antiquity.

After Baltruљaitis, there were four main names that were concurrently scrutinizing the concentric compositions in the 1940-1960s. The first one is Karl Lehmann Lehmann K. The Dome of Heaven // Art Bulletin.Vol. 27. 1945. P. 1-27. whose fundamental article devoted to the Roman imperial ceilings in the centered edifices is still relevant, even in spite of a range of new evidence that partially disproved his arguments This will be discussed in Chapter 2.. He discovered, using the 18th-century drawings of the ceilings' decorations as his evidence, and analyzing their iconography, the correspondence between the circular forms of Roman vaults with the ideas of power and celestial triumph. Helen Dow, for her part, was studying the origins of the Gothic rose windows Dow H.J. The Rose-Window // Journal of the Warburg and Courtauld Institutes. Vol. 20. No. 3/4. 1957. P. 248-297.. Her article is an example of truly exhaustive research. She traced the origins of the Gothic stained-glass windows back to the ancient Syrian oculi which had protective apotropaic powers and were associated with the solar signs. She also provided a truly thorough excursus to the history of a form of a wheel. Her article helped us to get a general sense of the broad spectrum into which fall plausible sources for the concentric schemes, and also supplied the tools in how to trace a history of a certain form, to which comparisons are relevant to make, and what the limits are of such an approach. The latter is evident in her attempts to deduce a mutual meaning between two objects by means of similarities between their compositions. If two objects share a form but not a context, it is next to impossible to make any kind of conclusions regarding their semantic relations.

The third researcher, namely E. Beer, also devoted two of her works to the origins of a rose window Beer E. Die Rose der Kathedrale von Lausanne. Bern: Benteli Verlag, 1952; Beer E. Nouvelles reflexions sur l'image du monde dans la Cathedrale de Lausanne // La Revue de l'Art.Vol. 10. 1970. P. 57-62.. Unlike her colleague, she studied not the rose windows in general, but one particular case, i.e. a quite unusual pattern in the composition of the Lausanne rose window. This window is characterized by a notably structured form. Beer suggested three sources for this concentric arrangement: cosmological, figurative, and ornamental works. Her research shows much more reliance on the contextual proximities between the studied examples, and thus her conclusions on the possible cosmological notions of the rose windows are highly persuasive.

Another scholar whose work falls within the same discourse but actually quite different is Harry Bober. Instead of researching the origins of compositions within the paradigm of religious iconography, he dedicated his to this day's extensively referenced article to one particular manuscript with scientific diagrams, namely the 12th-century scientific compendium with texts of Isidore, Bede and Abbo of Fleury Bober H. An Illustrated Medieval School-Book of Bede's "De Natura Rerum" // The Journal of the Walters Art Gallery. Vol. 19/20. 1956/1957. P. 64-97.. Bober treated the diagrammatic depictions in the manuscript not as mere scientific tools but as visual objects that have esthetic significance. This was a crucial step in terms of studying the relationships between diagrams and art because it equalized their status - the diagrams were elevated to the status that is closed of art. Interestingly, Bober doesn't say much about the esthetic values of the diagrams, but just his attention to the visual aspects of the manuscript in the collection of Walters Museum eventually resulted in a strong reevaluation of the role of the scientific images in art history.

In the 1980s, the article by M. Caviness was of particular importance. It was invaluable in our growing awareness of the liaison between the mode of thought and the art forms, and of the importance of a scheme in medieval visual culture Caviness M. Images of Divine Order and the Third Mode of Seeing // Gesta. Vol. 22. No. 2. 1983. P. 99-120. . She basically made a next step: she started to compare the complex compositions with the simpler structures of diagrams and came to the conclusion that both the diagrams and the religious compositions are made from the same modules. Certain works of H. Kessler, including his fundamental monograph on the Bibles from Tours Kessler H. The Illustrated Bibles from Tours. Princeton: Princeton University Press, 1977., examine the correlations between diagrams and religious subjects in a similar to Caviness' way which reflects a general tendency of this time period. But the work of Bianka Kьhnel Kьhnel B. Op. cit. stands out in her thorough attention to the context. She analyzes not only the immediate formal correlations between two genres, and not only the possible proximities in meanings but she places the works in a relatively broad context (in her case, it is the context of Carolingian intellectual culture), and asks whether or not the liaisons between science and religion were even possible in the given background.

Apart from the research in the area of medieval compositions, their origins and correlations within various media, other two groups of publications were relevant for our study. The first one is a research in medieval cosmology, geography and astronomy which has broadened the understanding of the correlations between scientific knowledge and the works of pictorial art in the medieval period. One of the important works, in this regard, is a publication on Carolingian astronomy and cosmology by Bruce Eastwood that facilitated our comprehension of a role that the ninth century played in the dissemination of science. Particularly useful was his short introduction to the interrelations between pictures and diagrams. The works of B. Obrist and E. Dekker are also helped to form our understanding of the scientific context.

The ornamental nature of the quincunx composition generated a need for a consultation with the theoretic works on ornament and its underpinnings. The dissemination of patterns from one medium to another were amply elucidated in a recent dissertation by Anna Bucheler Bucheler B. Ornament as Argument: Textile Pages and Textile Metaphors in Medieval German Manuscripts (800--1100): Ph. D. diss. Toronto: University of Toronto, 2014.

Hamburger J.F. The Place of Theology in Medieval Art History: Problems, Positions, Possibilities // The Mind's Eye. Art and Theological Argument in the Middle Ages / Ed. by J.F. Hamburger, A.-M. Bouche. - Princeton: Princeton University Press, 2006. - P. 11-32. . Apart from this latest publication, we were using a range of general works devoted to the meaning, importance, and psychology of ornaments, such as The Language of Ornament, and Ornament: A Modern Perspective by James Thrilling Trilling J. Ornament: A Modern Perspective. Seattle; London: University of Washington Press, 2003; Trilling J. The Language of Ornament. London: Thames&Hudson, 2001., and The Sense of Order by E. Gombrich Gombrich E. P. The Sense of Order. A Study of the Psychology of Decorative Art. Ithaca, New York: Cornell University Press, 1984..

After shortly discussing the present state of the question, we are now able to formulate the relevance and novelty of our research. We are primarily basing our methodology on Bianca Kьhnel's monograph but we are extending her approach beyond the boundaries of the Carolingian epoch and implementing it to the context of late antique and early Christian art. The origins of the quincunx scheme were not studied before. And yet this scheme, so ubiquitous at the turn of millennium, played a significant role in the formation of the compositional range of the Carolingian, Ottonian, and later Romanesque art. We are hoping that our work potentially may enrich the present state of research in terms of correlations between the subjects and their forms in early Christian and early medieval periods.

Chapter 1. The Quincunx Paradigm: Introduction to the Subject

This chapter is indented to be an introductory section to the three most basic questions regarding the quincunx composition. The first one considers the quincunx as a term, clarifying its usage in a historical perspective. The second one treats the earliest examples of the quincunx arrangement in Roman art and analyses the problem of its visual origins. And finally, the third entails the context of late antique and early medieval multi-medallion compositions. There, we will be dealing with the possible reasons for the dissemination of the quincunx scheme, particularly in a form of an ornament, leaving the figurative Christian imagery for the next two chapters. Before entering the area of subjects and meanings, one should understand the formative principals of the quincunx arrangements in their most abstract and general forms.

1.1 Quincunx as a word and as a term

Before undertaking the endeavor of researching the origins of five-medallion composition which in the modern historiography has received the name quincunx The word quincunx will be put in italics in all the cases we are referring to it as a Latin word used in Antiquity and Middle Ages. Otherwise, the italics won't be applied. , it will be appropriate to understand, in the first place, what this designation really means. More specifically, what it meant at the times when the quincunx composition initially emerged as a part of certain art forms. In order to do that, we will look at a group of antique and medieval authors and analyze their usage of the word quincunx in various contexts. The aim of the following overview is to comprehend how vary the meaning of this term now and in the Middle Ages is, whether it signified then what it does now, and whether the word had or had not an application to the scheme of five medallions that will be discussed in this paper. Such a clarification shall help us to implement the term correctly: only as a modern tool describing a particular composition, or as a phenomenon that had an existing equivalent in the studied time period.

The word quincunx consists of two parts - quinque meaning “five” and uncia meaning “one twelfth”, which explains its overall definition as five-twelfths. The word was used in the name of the Roman bronze coin issued at the end of the third century BC, in circa 211-208 BC Sear D.R. Roman Coins and Their Values. Vol. 1.: The Republic and The Twelve Caesars 280 BC - 96 AD. London: Spink, 2000. P. 219., with the value of five-twelfths of an as (fig.1). Although not always arranged in a checkered manner, the pattern of five pellets was frequently a part of the coins' design, an indication of their cost. The monetary production of the quincunxes stopped right after the Second Punic War but the word itself survived, and came to be regarded as the identification of this gauge - five-twelfths - and also of the pattern Ernout A., Meillet A. Dictionnaire etymologique de la langue Latine. Paris: Klicksieck, 2001. P. 557. . In fact, the coin is most certainly not the first precedent of the usage of this numerical fraction, but it definitely attests to its existence and its connection to the name. Multiple authors in Antiquity, such as Marcus Terentius Varro, Caesar, Cicero, Columella, referred to the term quincunx to denote both the configuration of five and the measure Vaan M. de. Etymological Dictionary of Latin and the other Italic Languages. Leiden, 2008. P. 509; Walde A., Hoffmann J.B. Lateinisches Etymologisches Wцrterburch. Heidelberg,1938. P. 345. . Particularly, Varro in Res Rusticae implements it to define the very specific arrangement of trees, planted in a checkered order to achieve a perfect distance for an optimal harvest. Columella's reasonings in De Re Rustica are similar to the ones of Varro's: he recommends the quincunx planting in the vineyards with fat soil as the best way to make a good distance between the trees in order to give their branches a space to spread Lucius Junius Moderatus Columella. On agriculture. Vol. I: Res Rustica / Transl. by H. Boyd. Cambridge, Massachusetts: Harvard University Press, 1941. P. 320-321. . And even Cicero mensions the beaty of the quincunx plantations in his portray of an estate of a Persian prince Cyrus and his hospitality towards his Spartan guest Lysander. Manifestly, this disposition of trees was such a pleasure for the eyes and so well for the healthy growth of trees that Quintilian uses it as an analogy and metaphor for a true eloquence which should be both rhetorically delightful and produce a strong emotional response.

Another instance of the application of this word we encounter in Caesar's De Bello Gallico where he describes a complex fortifications consisted of five rows organized in a quincunx C. Julius Caesar. De Bello Gallico / Transl. by W.A. McDevitte, W. S. Bohn. New York: Harper & Brothers, 1869. §7.73.5. . The quincunx disposition is supposed to provide a strong and stable defense from the adversaries and constitutes a sort of a trap. The plantation and fortification examples are equally important considering they both testify for the usage of a quincunx in the description of visual structures, which is particularly relevant for our subject matter.

Horatius attests to the quincunx as a part of Roman arithmetical system: in one of the scenes of his De arte poetica, boys are learning how to count fractions by dividing an as Q. Horatius Flaccus. Opera / Ed. by F. Klinger. Berlin: De Gruyter, 2008. P.323.

“Dicat filius Albini: si de quincunce remota est

uncia, quid superat? poteras dixisse.” “triens.”. There the quincunx minus ounce results in one-third. It proves that, indeed, its value equals five-twelfths, and, more importantly, that the quincunx in the first century BC is routinely used as a measure of a fraction. In the first century AD, the poet Persius employs the initial meaning of the quincunx, in a sense of a currency of a modest value. But the way he treats the word is closer to the abstract number than to the value of a real coin “quid petis? ut nummi, quos hic quincunce modesto

nutrieras, pergant auidos sudare deunces?”. (exactly this meaning will be playing the most significant role in the medieval understanding of a word). Gradually the quincunx acquired this multi-purposed sense of measurement - still maintainig the value of five-twelfths, of course, but disconnected from the very specific context of coinage. For instance, Frontinus estimates the capacity of the pipes served for the water supply in Roman aqueducts in quinaria and quincunces. Martial in the XXVII epigram of the Book I uses this word instead of a half-pint or a small glass to assess how much he drank the night before Martial. Epigrams / Transl. by W.C.A. Ker. London: William Heinemann, 1947. Vol. 1. P. 46-47. , and in another poem, he mentions `five measures', or quincunces, that was mixed for a guest's drink Ibid. P. 110-111.. Marcus Cornelius Fronto in his letter to Antonius Pius indicates the size of an inherited portion of an estate as five-twelfths, i.e. quincunx The Complete Works of Marcus Cornelius Fronto / Transl. by C.R. Haines. Hastings: Delphi Classics, 2018. P. 831. . That is to say, the quincunx in all those examples is a perfect demonstration of the ingrained in all spheres of Roman life metrology, the knowledge of measures, lengths, volumes, and also of money. In Late Antiquity and Early Christianity, it is much easier to confront the meaning of the quincunx as a number compare to its association with a visual structure of five dots. Priscianus mentions it in De figuris numerorum treating the word in the context of mathematical terminology. Ausonius in his Eclogues, specifically in De Ratione Librae, uses the word quincunx as a five-twelfths measurement, though not in its purely technical aspect but rather placing it in the context of poetical and philosophical collations “Si defuit uncia, totus / non erit as nomenque deunx iam cassus habebit. / nee dextans retinet nomen sextaiite remote / iam quincunx tibi nullus erit, si gramma revellas”, which reads: “If one ounce is wanting, it will no longer be a pound, but being short in weight will be called deunx (eleven-twelfhs). The dextans (five-sixths), too, does not retain that name if a sextans (one-sixths) be taken from it… Take away one scruple, and you will have no quincunx left you”. See: Ausonius. Works / Trans. by H.G.E. White. London: William Heinemann, 1919. Vol.1. 174-177. . Being on the verge between specific and abstract, the usage of the term quincunx in this description nevertheless remains within the subject of measurments, specifically the measurements of weights. The author doesn't refer to any visual structure, as the case with the earlier examples of antique writes, and doesn't offer any connection between the quincunx as a measure and as a pattern.

Following the same tradition, Bede continues to use this word in the context of weight measurements in De ponderibus, meticulously clarifying the nuances of the ounces as portions of a pound However, there are some variations in the word choice from manuscript to manuscript: it seems that scribes used the forms quincunx, quincuus and cingus interchangeably. Indeed, in the fourth chapter of De temporum ratione, devoted to the calculation of duodecimal fractions, Bede again reiterates the substitutability of these words: “Quincunx, sive quincuus, quinque uncis”. Furthermore, the fourth form of the word is introduced in the treatise De arithmeticis numeris, allegedly also attributed to Bede, where the author suggests a variant quinquus for the same measure of five ounces. Apparently, the classical Latin didn't know the forms quincuus, quinquus and cingus. See: Poetarum Latinorum Medii Aevi / Herausgegeben von K. Strecker.; Monumenta Germaniae Historica. Weimar: Hermann Bohlaus Nachfolger, 1951. Tomus VI. P. 189; Venerabilis Beda. Opera Omnia / Beda Venerabilis; ed. by J.-P. Migne; Patrologiae Cursus Completus. - Paris, 1862. Vol. 90. P. 307. . A novelty in Bede's usage of the word is his application of the quincunx to the context of the time reckoning. The author accentuates that it is “no unworthy” to know how to calculate them properly “for one can apply it to the calculation not only of coinage but of time and of other things” Bede. The Reckoning of Time. Bede. The Reckoning of Time / Bede; transl., with introd., notes and com. by F. Wallis. Liverpool: Liverpool University Press, 1999. P. 16.. Partially drawing on Bede's works, Rabanus Maurus' De computo offers similar treatment of the term and places it in the same framework. In the 10th century, the quincunx as a number also appears in the musical context, in the theoretical treatise on Gregorian chant by the Cluniac monk Odo Sancti Odonis Abbatis Cluniacensis Secundi Opera Omnia / Ed. by J.-P. Migne; Patrologiae Cursus Completus. Paris, 1881. Vol. 133. P. 810.. In the 11th century, Hermann of Reichenau in his treatise Prognostica de defectu solis et lunae implements the term to measure time which, after Bede, was apparently a usual practice He criticizes the inaccuracy of the existing units, including the quincunx, suggesting a more fractional and precise one - a portiuncula, a 127th of an hour. See: Germann N. De temporum ratione. Quadrivium und Gotteserkenntnis am Beispiel Abbos von Fleury und Hermanns von Reichenau. Leiden; Boston: Brill, 2006. P. 224, 342..

As is obvious from all these examples, the most frequent implementation of the word quincunx in the Middle Ages was a numerical one - an explicit continuation of the antique tradition. By the contrast to Roman authors, the medieval texts starting with Bede implement the term to the scientific and musical context. The significant difference between the antique and medieval usage of the word occurs, strictly speaking, not in the meaning itself but in the ways this meaning is treated. In Antiquity, its application was a part of day-to-day life, and designated both the pattern and the number, appearing in the contexts of coinage, agriculture, mathematics, law, war, etc. In the Middle Ages, this system of measurements was gradually winding down and disappearing from the quotidian, leaving its traces only in scientific treatises.

But the question arises whether or not the word quincunx was associated with the measurements exclusively, or with the pattern as well? Presumably, the measure and the pattern were so closely linked to each other that, in order to save time, the sign of five dots organized in a centered blueprint was used in medieval texts instead of a word quincunx Breal M. Dictionnaire Etymologique Latin / M. Breal, A. Bailly. Paris: Librairie Hachette, 1906. P. 301. . This fact is mentioned in one of the etymological dictionaries we were consulting with, but the text doesn't provide references to any specific manuscript, and the consequent investigation didn't give any positive results either. On the contrary, in the realm of Anglo-Saxon manuscripts, particularly in Bede's works, the word quincunx is presented by a completely different sign that has nothing to do with the antique five dots of a dice (fig.2) The examples include: Karlsruhe, Badische Landesbibliothek, MS Aug. 167, 9^th century, f.25r; Oxford, St. Jones College, MS 17, f.68r; London, British Library, Royal MS 15 B XIX, f.45r. . That said, there is a close analogy for the quincunx sign - a cross of dots that might be seen in a number of early medieval manuscripts devoted to various subjects (fig.3). And yet the sign is inconsistent in terms of the frequency of its appearance, and of the exact places he accompanies. For that reason, its specific meaning is hard to define The crosses in general are vastly recurrent in the manuscript annotation, and their purpose might vary from the text structuring to correction, annotation and dialogue marking. See: Steinova E. Notam Superponere Studui: The use of technical signs in the early Middle Ages: Ph.D. diss. Utrecht: Universiteit Utrecht, 2016. P. 215, 239, 273.. Therefore, the connection between the word quincunx and its visual representation of five dots in the Middle Ages remains doubtful. Furthermore, no treatise accompanied by the five-medallion diagram implies any connection between the images and the word quincunx. Was the quincunx ever applied to the context of religious imagery? Did medieval thinkers discuss images such as Maiestas Domini in these terms? Apart from this idea being somehow outrages, we also don't have evidence for that.

The earliest post-antique mention of a visual representation of the quincunx which is at least partially reminiscent of an antique sign dates only to the middle of the 17th century Latham R.E. Revised Medieval Latin Word-List from British and Irish Sources. London: The British academy, 1983. P. 389. . Notably, Henry Spelman in his treatise Aspilogia on coats of arms applies the word quincunx to the description of heraldry Spelmanni H. Aspilogia // Nicolai Uptoni. De studio militari. Libri Quatuor; Iohan de Bado Aureo. Tractatus de Armis; Henrici Spelmanni. Aspilogia. Ed. by Edoardus Bessaeus. London: Typis Rogeri Norton, 1654. P. 139: “Geftantur autem lilia et nonnulli flores, alii quincunci ordine per aream totam dissipata, quod ideo quincuncem apellamus, vel divaricationem”.. This meaning was apparently maintained during the medieval era, or otherwise, it wouldn't have appeared again in the early modern period. By this time, the numerological notion of the quincunx and its association with measurements, if not forgotten completely, played a secondary role since in its modern usage it retained only one connotation, namely with the discussed above sign of five dots or a five-medallion structure.

Now, this term might be understood slightly better. The historical overview was needed for couple of purposes. First is to show that the quincunx paradigm has Roman roots not only in terms of the applied to art visual structure, which will be discussed below, but also philologically. By and large, if something has a name, it exists. Albeit there is no indication for a purposeful employment of the word to the context of imagery, it is still an important testimony for Romans having a descriptive term for a pattern we are analyzing, even if it appears in different contexts. Second is to distinguish between the Roman, medieval, and modern implementations of the word. Although the quincunx had visual connotations in Roman tradition, they apparently very much disappeared during medieval times. And thus, referring to this word, we are recalling only its modern sense, namely the composition of five dots, or five medallions of all shapes and sizes but mainly circular, in ornaments, scientific diagrams, and religious iconography. Interestingly, in modern English language, the term was typically used to denote patterns, or the five-medallion diagrams in the works on medieval cosmology and computistical science For example: Germann N. Op. cit; Eastwood B.S. Ordering the Heavens. Roman Astronomy and Cosmology in the Carolingian Renaissance. Leiden; Boston: Brill, 2007., but it is only lately that the quincunx was actively introduced to the art historical domain of religious iconography in the fundamental work of Bianca Kьhnel on Carolingian intersections between art and science Kьhnel B. Op. cit.. Hence, this monograph is our primary point of reference for the correct implementation of this word.

1.2 The Roman origins

When the quincunx appeared in art and why? Apart from clarifying the terminology, the previous segment of our introduction to the subject showed that in the ancient Rome, the idea of the quincunx disposition, at least in the landscape design, and according to Quintilian, was esthetically pleasing. Although we do not have such accounts for the art forms, it was in the context of Roman Empire where the quincunx composition emerged for the first time. Conceivably, its appearance might be connected to the centered forms of Roman edifices. The fact remains that the initial episodes of the quincunx compositions - at least those which survived in some form - belongs to the media of ceiling decorations, specifically in two centered buildings of the 1st and 2nd centuries AD (fig.4-5). Later episodes demonstrate that the quincunx pattern was ubiquitous in the floor mosaic medium Our relatively meticulous search through the compositions of Pompeian frescoes didn't give any results, but perhaps a more exhaustive investigation is needed. , primarily in a format of a repeated checkered ornament (fig.6).

It is tempting to suggest that the practical reasons for organizing a space by means of the quincunx composition came first, and then the same pattern was implemented elsewhere On the psychology of the construction of patterns and compositions, see: Gombrich E. P. The Sense of Order. A Study of the Psychology of Decorative Art. Ithaca, New York: Cornell University Press, 1984. P. 1-17; Trilling J. Ornament: A Modern Perspective. Seattle; London: University of Washington Press, 2003. P.21-71.. By `practical reasons' we mean that the structure of the surface that is supposed to be decorated suggests, in some way or another, a set of formal variations for a future composition. Inter alia, the vaults and ceilings of the centered buildings impel to highlight the middle section, in comparison to the pavements where the possibilities of decorations are endless. However, the case is more complex, and actual direction of influences is hard to define. An alternative hypothesis that the checkerboard pattern came from the pavements and then was adopted for the centered compositions in the ceiling decorations is also feasible. In that way, it emerged as a more simple shape that was created by multiplication of one element and its repetition in the quincunx order, and then transformed from a pattern to a composition with hierarchical structure by means of centered architectural form it ended up occupying. Perhaps the solution is to address two distinctive quincunxes, both of which - quincunx as a pattern and as a hierarchical composition - eventually came to the pavement decoration and might be scrutinized for an instructive comparison. The first one, like in the 6th-century pavement from Roman Teurnia (fig.6), constitutes a small unit of a checkered pattern and probably originated on the large free surfaces of floor mosaics The vast floor surfaces also entail some regularities such as the propensity to arrangements in sequences. They might be organized in three contrasting ways, depending on what is chosen as their structuring unit. It might be a small integral module which is repeated monotonously or with any sort of variations; or it could be a more complex compartment which constitutes an independent composition and which is also exponentially repeated; or the whole area may be organized as one compositional unit. Regarding the quincunx, it falls into the second category.. The second one, such as the 4th-century marble mosaic from Ostia (fig.7), seems too complex and structured for being a result of a sporadic, externally unprovoked appearance. Its middle roundel is much bigger than the lateral ones and additionally comprises a rhombus, a square, and a smaller circle. Lacking of any subject, the pattern still expresses a strong sense of hierarchy between the center and periphery. This type of the quincunx, indeed, might have been stimulated by the centered form of Roman ceilings. The combination of a circle within a square space generates a need for four additional elements at the square's corners, shaping the quincunx pattern. Hence, the quincunx is one of the ways to interpret the juxtaposition of the circle within the square, which constitutes a common dilemma of centered architecture where a dome placed above a square compartment produces a problem of a conversion between angular and circular forms. In the ornamental decoration of a ceiling, such a difficulty of harmonizing the central and peripheral configurations might be resolved by the quincunx scheme. Indirect evidence for such an origin of this type of the quincunx is that, in the early Christian context, it appears much regularly in the vault and ceiling decorations of the centered chambers than in the mosaic pavements where the chess-like pattern is prevalent The examples are: Galla Placidia mausoleum, Ravenna; baptistery of St Giovanni in Fonte, Naples (fig.36-37); the vaults in the Catacombs of Domitilla and Catacombs of Marcellinus and Peter, Rome (fig.15-16). .

We are particularly interested in this second type of the quincunx, organized more complex and hierarchically, which will underwent an extensive development in the later Christian context. Its earliest example in the Roman pagan tradition occurs in the 1st century AD decoration of a vault of Nero's Domus Aurea (fig.4). The work itself was not preserved, but there is a surviving 18th-century drawing showing Jupiter placed in a circular medallion at the center and accompanied by four symmetrically arranged oculi opening into the sky Lehmann K. Op. cit. P. 10. Fig.27; Dow H.J. Op. cit. P. 252. . At Hadrian's villa, the ceiling is allegedly contained, as per the survived engraving (fig.5), a medallion with a solar god in the center framed by the circles with zodiac signs arranged not in a circular manner but much like a quincunx, though more complicated than the usual one Mathews T. The Clash of Gods: A Reinterpretation of Early Christian Art. Princeton: Princeton University Press, 1999. P. 145. . According to H. Joyce Joyce H. Hadrian's Villa and the “Dome of Heaven” // Mitteilungen des Deutschen Archaologischen Instituts, Romaische Abeteilung. Vol.97.1990. P. 347-375. , those engravings are not a reliable source for the assessment of Roman art in terms of subjects and style since they are copies of copies and have a feeble connection to the original monuments. However, T. Mathews fairly points out that with respect to overall structure and composition those engravings reproduce authentic Roman practices of framing the ceilings Mathews T. Op. cit. P. 148. . Therefore, although the first century evidence for the subjects in the quincunx arrangement is lacking, we know that this composition was applied to the architectural decoration as early as the time of the emperor Nero. The earliest survived example of the quincunx dates to the 2nd-3rd centuries and constitutes the short side of a lead sarcophagus where a central medallion is occupied by dancing Eros, combined with four medallions comprising female busts at the angles of the square composition (fig.8). This structure remarkably corresponds to what we will see later in religious imagery, namely in the iconography of the 9th-century Maiestas Domini. Christ will replace a cupid in a central roundel, and instead of busts of the deceased the lateral medallions will be occupied by the four evangelists. And even a lozenge of the lead sarcophagus composition will take its rightful place in these later Christian instances (fig.91).

In spite of its Roman origin, the quincunx scheme is a relatively rare occurrence in the late antique context. These three examples of which only one survived are the main testimonies of the quincunx composition during the 1st-3rd centuries AD. Hence, it is near impossible to study the earlier quincunx tradition in terms of its subject-matter. Moreover, for any sort of comprehensive analysis, including formal, we simply need more data.

1.3 Quincunx in the context of multi-medallion compositions

Now, we are leaving the context of late antique Roman art and entering a broader perspective of multi-medallion compositions in early Christian and early medieval traditions. Since the preserved quincunx precedents in the first centuries AD are insufficient with an eye to gathering any sort of statistics, one shall expand the searched timeframe. Moreover, here we are not aiming exclusively for the quincunx patterns but pursuing a more general context of early Christian and early medieval multi-medallion compositions. The aim is to identify the instances where the quincunx appears: in what kinds of media the quincunx occurs, and the proximity to which forms might be regarded as the catalysts to its emergence. The key interest of the following overview is essentially formal, hence the figurative Christian iconography and its influence on the quincunx composition will be regarded further in the second and third chapters.

In late antique, early Christian and early medieval periods, the medallion compositions exist in a variety of forms. They might emerge in the mosaic medium in circular arrangements, like in the 4th-century example from Tunisia (fig.9), or create chains of roundels as in the 4th-century ceiling of the Santa Costanza mausoleum (fig.10), or just regularly cover the surface, as in a pavement from Petra (fig.11), or even form an arch (fig.12). In the book illumination, the range of medallion compositions is also quite wide. They may fill the arch, as in the Golden Canon Tables from Constantinople (fig.13), or being a part of circular composition, as in the Rossano Gosples (fig.14), etc. There is no need to list here all the media and all the examples of possible variations, suffice it to say, there are quite a lot of them. Our question is in what circumstances they form the quincunx?

After having analyzed hundreds of medallion compositions, we are able to identify two main situations in which the quincunx might emerge. It doesn't mean that it appears only under these two conditions exclusively, but these are the cases which demonstrate a relative regularity. The first one is a combination of a circle and a square, which we already partially discussed in relation to the Roman origins of the quincunx composition. Indeed, this type of the formation of the quincunx often occurs within the decorations of the centered ceilings. It is particularly common in the context of catacomb paintings. The vault of Cubiculum 3 in the Catacombs of Marcellinus and Peter (fig.15) is one of the most illustrative examples of such because it constitutes an elaborate case in which the quincunx is only a small part of a bigger arrangement, and the interplay between a circle and a square is so evident. Here, the quincunx is created by the rectangular in compositional kernel and by the four roundels attached to its angles. They are needed to smoothing the transition from angular center to a bigger circle, which comprises all the mentioned elements. In the Cubiculum 78, a complex concentric composition fills the ceiling of a square chamber, and in order to mutually harmonize both forms, four additional roundels were added to the corners of the ceiling (fig.16). Thus, the quincunx in both of these instances functions as a strategy to balance the juxtaposition of a circle and a square.

The second circumstance in which this composition appears is the presence of a cross form On the cross pattern, see: Gombrich E. Op. cit. P. 247-249. Beatrice Kitzinger's recent book on the iconography of the crosses and their symbolism in the Carolingian period was, unfortunately, out of our reach, see: Kitzinger B. The Cross, the Gospels, and the Work of Art in the Carolingian Age. Cambridge: Cambridge University Press, 2019. (non vidi); Bianca Kьhnel was the first to connect the cross and the quincunx forms as applied to the context of Carolingian book illumination, see: Kьhnel B. Op. cit. P. 55-52. . Multiple cases in various media are the testimony for that. Usually, the construction of the quincunx starts when a cross has a medallion at its center - in this situation four additional ones emerge between the arms of the cross. This is not the only way in which the cross provokes the appearance of the quincunx, though it is the most common one. For a start, one may look at some of the mosaic pavements both in pagan and Christian contexts. In the mosaics from villa Piazza Armerina in Sicily (fig.17) with the entirely pagan subjects as its theme, namely an erotic scene in the middle accompanied by female personifications, the cross is formed by four hexagons and four half-medallions, and the set of other four roundels compose the quincunx. Here, the quincunx and the cross are relatively equal, but in the Christian mosaics the cross starts to prevail. For instance, the 8th-century pavement from Jordan (fig.18) comprises a composition with five big medallions designed as a cross, and four smaller and flatter ones between them which make up the quincunx.

That the latter is a stepbrother of a cross should be evident from the testimony of jewelry, coins, and other metalwork. One can take as a witness the obverse and reverse of an 8th-century Anglo-Saxon coin issued in the rule of Kentish king Offa of Mercia (fig.19) The Making of England: Anglo-Saxon Art and Culture AD 600-900 / Ed. by L. Webster, J. Backhouse. London: British Museum Press, 1991. P. 249. No.218c. . What interesting is that the center of its reverse incorporates five pellets in a form of a cross, and an obverse accommodates the quincunx, which becomes a sort of a mirror reflection of the cross, and its immutable companion. A 9th-century copper coin from York, though of a lesser quality, demonstrates similar principle: the cross and quincunx are placed on the reverse and obverse respectively (fig.20) The Making of England. P. 155-156. No.122. . These two forms become two sides of a coin, in a very literal sense of the expression.