Thea Burns. 'Compositiones variae'. A Late Eight-century Craftsman's Technical Treatise Reconsidered

Translation by Francesco Mazzaferro
CLICK HERE FOR ITALIAN VERSION

Thea Burns
Compositiones variae. 
A Late Eight-century Craftsman’s Technical Treatise Reconsidered
London, Archetype Publications, 2017, 224 pages

Review by Giovanni Mazzaferro

The Compositiones variae (or Compositiones ad tingenda musiva or Compositiones lucenses) are a set of recipes on art techniques, preserved within the Code 490 in Lucca’s Biblioteca Capitolare Feliniana, the Capitulary Library created immediately after the death of Felinus Sandeus (1444–1503), the bishop of the Tuscan town of Lucca, on the basis of his will. It is one of the most famous recipe books of this type. The many experts, who studied it since the French scholar Jean Mabillon rediscovered the text in 1686, agreed that it was written around the end of the 700 AD.

It should be immediately said Thea Burns’ volume is not offering a new comment or a new critical edition of the Compositions. Therefore, when you have her book at hand, you can read neither the Latin original nor a modern translation of it. As for a recent Italian version, I should like to refer to the edition published by Adriano Caffaro in 2003 [1]. This was a definite methodological choice by Burns: when we are faced with a manuscript - the authoress wrote - there are two types of information reaching us: on the one hand, the content of the text, and on the other hand, the object (the manuscript) in its own physicality. A scientific examination of the latter allows us to consider its origins and history. Manuscripts are, by definition, unique objects, even when they are copies of other texts. It is only our modern idea of authorship of a work that drives us to search for its original version, but this does not exclude that each of the stages in which the tradition of a given text has been performed has its precise meaning.

In the specific case, the origin of the recipes contained in the Compositiones variae is unclear, due to the lack of other witnesses on the subject. There are those who saw a link between the Compositiones and the Egyptian papyruses of Leiden and Stockholm; there are others who considered the manuscript as a very corrupt version informing the reader about the so-called 'first historical alchemy' (see in this blog the review of Mappae clavicula). i.e. of the technological heritage on metal processing in Egypt and then passed from there in the Greek world. According to many, this know-how (at least in its technical value and not in the 'philosophical' meaning attributed to alchemy in the Renaissance) came to Italy on the end of the parable of the Roman Empire. The Compositiones variae would be an example thereof (and some features are unquestionable: for example, one recipe of the Compositiones is the transliteration in Latin characters of a Greek original. See page 2). Some went even further, and highlighted the similarity of the Compositiones with cuneiform inscriptions on some Assyrian tablets found in Nineveh, speaking about making glass, metals and pigments.

Most likely - the authoress is right - we will never be able to recompose the general picture and to outline the 'family tree' of the Compositiones. But even if we were able to do so, the point emphasized by the authoress is different: focusing all attention on the historical reconstruction and the chronology of the work would devoid the attention due to the individual stages that make up its physicality and to what they are able to tell us about the society in which they were processed, starting with the examination of the code hosting the recipes. 

The Good Sheperd and incipit. Biblioteca Capitolare Feliniana, Codex 490, folio 348r (Genealogiae totius bibliothecae)
Source: http://www.artearti.net/magazine/articolo/a_lucca_in_compagnia_del_professor_ragghianti/

The Codex 490 and Lucca

The Codex 490 of Lucca’s Capitular Library is actually a miscellaneous codex hosting dozens of complete texts (or fragments thereof), including some which are more or less known, such as the Liber Pontificalis and the Etymologies by Isidore of Seville. As an element of reflection, I would like to emphasize that the codex, for the very fact of existing, has created a problem for the scholars: that of the transmission of culture in the so-called darkest centuries of the Middle Ages. According to tradition, this circulation would be exclusively limited to monasteries. In this case, instead, the text circulated outside of this reality, so much so that it is assumed that in Lucca existed either a secular 'scriptorium' (i.e. a transcription centre) or even a "school" of writing, which must have survived in some way to the disintegration of the Roman empire and the barbarian invasions. In reality, there was however a common element with monastery culture: at least in its most significant units, all texts of the codex had a religious theme. The Compositiones were the exception; in the Codex (consisting of 355 folios), they ranged from the folio 217r to the 231r. [2]. In fact, however, if we look at Lucca's historical situation in that period (see chapter three) and the artistic and craftsmanship of the city (see chapter six), the presence of the Compositiones was indeed coherent, because it referred to the interest of the local bishopric for the realization of artistic artefacts to enrich ecclesiastical circles.

Without discussing ancient Etruscan and Roman origins, it is worth pointing out the importance of the city of Lucca in the Langobardic time. Lucca was the capital of the Langobardic Duchy of Tuscany, situated in a strategic position along the Via Francigena, which connected Rome to Pavia (the Langobardic capital) and more generally to the rest of Europe. Above all, it was a stage for pilgrims to travel from and towards Rome and presumably a commercial centre, even on a small scale. In 773 the city fell under the rule of Charlemagne and the Carolingians. It was in this historical reality, that is in the early Carolingian age, that the Codex 490 or, at least, the units constituting it (and probably originally separated from one another) were created in Lucca. Contrary to what one might think, the transition from the Langobardic to the Carolingian rule must not have been a revolution. It would seem so, at least in the ecclesiastical world. The bishop of the city, the Langobardic Peredeus, was brought to France and (so to say) 're-educated' to the new needs of Charlemagne’s world; thereafter, he returned to Lucca and kept his office until his death. The figure of the bishop was not only religious. Most likely, the bishop was also the main landowner in the area and played a non-indifferent political role at the local level. At his death (780), Giovanni and Jacopo, both sons of the Langobardic Teutpertus, followed him. If a transfer of power occurred between Langobards and Carolingians, this was done only gradually, without the immediate substitution and dispossession of the most important families of the defeated rulers. Naturally, the price to pay for those who remained in positions of responsibility was that they were forced to identify themselves with the winner's policy.

It is known that in 789 Charlemagne emanated the famous 'Admonitio generalis', which advocated a profound reform of the ecclesiastical world and of the educational model. From a religious point of view, this meant a gradual nearing to papal authority. The educational model, on the other hand, proposed the creation of schools that were to be hosted by cathedrals and in each monastery, in which to undertake a substantial process of spreading literacy (fundamentally, dedicated to nobles and clergy). Precise rules for writing were coded; the clarity of the texts and their readability was recommended; standardizing written language meant, first of all, making more efficient a system governed by universal standards.

We do not know exactly when Carolingian reforms began to deliver their fruits in Lucca, but one thing is certain. The Compositiones (or, better, the texts composing the Codex 490) were not part of this design and were indeed a testimony of a writing full of errors, but especially of changing conventions depending on who was responsible for writing (at least one dozen of scribes). It appears correct, in short, to place Codex 490 as the last result of a rapidly changing system, which would eventually be replaced by new (and simpler) rules.

The fact itself that the codex as a whole did not have an index (or a table of materials), as well as any numbering distinguishing the various writings that compose it, suggests (as we will see below) that the writings it contained had an individual life before being bound in the volume that has come to us. 

Workship from Lucca (Langobardic epoch). Pluteus with a fragment of a cross and animals, about 600-620 a.C.
Source: Sailko (Francesco Bini) via Wikimedia Commons

Physical evidence from Codex 490 and a hypothesis about its formation 

The physical examination of Codex 490 is facilitated by the fact that the folios composing it are nowadays kept loose and that the binding in which they were united is still preserved. The first and the last folios were glued directly to the volume leather cover. This is why they have a brown colour that is due to prolonged contact with the leather itself. But the careful examination of the sample shows that these two folios were not part of the code originally. If one pays attention to the holes left by the beetles on the cover, it is easy to see that they match similar holes in the current second and penultimate page of the code, but not in the first and the last ones. Moreover, the second and the penultimate pages have text stains and discolorations, which are clearly due to prolonged contact with a cover.

One thing to consider, however, is that these features also appear in many of the copybooks that make up the first or last page of individual works contained in the code, a clear indication that these copybooks were kept loose and unprotected for a long time, and therefore had an autonomous life. The technical term used to describe them is the Latin libellus. It is also logical to think that, for short liturgical writings, it must have been more functional for their use to keep them loose (and even folded and carried in bags or pockets). Gathering them in a single volume meant to take care of their conservation in the long run. It is not certain that such 'historical' awareness (or just curiosity) occurred immediately.

For example, the authoress advanced the following hypothesis about the genesis of this code: "I propose [...] that when the information in the individual texts today making up Codex 490 became dated, probably at some point during the ecclesiastical reforms carried out under Charlemagne, the information they contained was replaced by other, more up-to-date texts which the Lucca clergy then consulted in preference. In addition, writing practices and language may have changed to such a degree that reading became impractical or impossible. In this scenario, the unbound contents of the present Codex 490, still in individual book and fascicle format, were stored away in a chest, a cabinet or on a shelf and forgotten. [...] At same point, perhaps during the reforms of the fifteenth century [3] [...], and certainly before Mabillon's visit to Lucca in 1686, the multiple loose booklets were collected together and sewn onto the cords for protection and safekeeping. The cords, remnants of which are extant today, were then or later inserted into wood boards; it is not known when the full leather cover and clasps, still preserved, were added to the boards" (pp. 76-78).

Workship from Lucca (Langobardic epoch). Fragment of a slab with a cross and flowers, about 600-650 a.C.
Source: Sailko (Francesco Bini) via Wikimedia Commons

Dating

If this assumption was true, the question of dating the code would become much more complicated (but perhaps less important). It is clear that the individual units were written in different times, perhaps not so close to each other as criticism was induced to think, taking the view that the physical creation of the codex itself was also to be traced back to the end of the eighth century or the beginning of the following one.

Normally (for the presence of dates in some texts, but also on a palaeographic basis), it is believed that the writing of the texts should date back between 787 and 816 after Christ. However, the authoress believes that it would be a methodological mistake to attribute the date of writing of one text to all others, and I cannot but agree with her. What is not being said, but I personally could not help but think, is that if Burns's reconstruction was true, and if the constitutive units of the Codex were texts that suddenly became 'old' after Carolingian reform, that reform begun - as seen - in 789. Even though it is entirely logical to think that the new rules took a few years before establishing themselves in Lucca, it would be intuitive to believe that most of the writings could be a few decades earlier than it has been believed until now.

Golden cross of Langobardic epoch, (from Lucca), VI-VII sec. a.C.
Source: Sailko via Wikimedia Commons

A non-literary text?

Burns's work offers a simple, but at the same time fascinating, work hypothesis. After all, it is entirely logical to think that the true 'fortune' of the texts in Codex 490 was that they have been set aside and not used for a long time (thus avoiding damage) and then rediscovered much later on.

And yet, I would not be sincere, if I said that her methodological approach convinced me fully. In my opinion, the study of the physical object must be complementary to the palaeographic and philological analysis of the individual texts. I hope that Burns will have no trouble agreeing with me, and that her work will soon become part of an ideal diptych that needs to be enriched by a targeted work on the texts.

Nevertheless, there is still something that did not convince me: in two occasions (p.1 and p. 81), when speaking of the Compositiones variae, the author defined them as "a non-literary text". Here, on that aspect, I'm afraid I cannot agree. The Compositiones were, in all respects, an integral part of a literary genre (the literature of recipes), with its rules, rhetoric and transmission mode. The fact that, in this particular case, the witness that has come to us is particularly corrupt does not mean that we can consider it as a ‘non-literature’ text. I am referring in particular to the proceedings of the seminar “Trattati e ricettari per colori. Una metodologia di studio nell’ambito delle scienze umanistiche, Milano 6 dicembre 2013" (Treatises and recipe-books for colours. A methodology of study in the field of humanities, Milan, December 6, 2013), which are today available for free on the number 16/2016 of Memofonte Studies. In particular, I am quoting below what Sandra Baroni and Paola Travaglio wrote in their ‘methodological premise’ on the lack of research of the texts relating to the tradition of a manuscript (18): "Not to look for these texts led to the mistaken belief that the Compositiones lucenses were the "monument" of art treatises in the Middle Ages, «the oldest among the manuals that the High Middle Ages dedicated to art making», while they are only the precious but - for transmission damages – uncoherent testimony of the Ancient Latin translation of an originally perfectly organized and coherent Hellenistic work. In the late-antique world this text existed first in Greek and later in Latin, and the text remained almost intact at least until the Carolingian reform, when problems of consecutio [note of translator: sequencing] began to emerge. Those problems could not be easily restored, and later were increasingly aggravated during the tradition." It is not to be excluded that scholars will not manage to formulate a convincing hypothesis about how the work was originally structured. In some cases, there have been some findings (four recipes in the Compositiones, for example, are virtually identical to the same prescriptions of Mappae Clavicula [4]). And, on the other hand, it is sacrosanct to start from the physical features to reason on the material birth of the work and also on the environment that produced it. As long as – however – one does not forget that the two things are intimately linked to each other, and both can enrich our wealth of knowledge on technical literature, a particularly intricate, but definitely fascinating genre.

NOTES

[1] Adriano Caffaro, Scrivere in oro. Ricettari medievali d’arte e artigianato (secoli IX-XI). Codici di Lucca e Ivrea (Writing in gold. Medieval recipe books on art and crafts (IX-XI centuries). The Codes of Lucca and Ivrea), Naples, Liguori Publishers, 2003. It should be noted that the edition did not always enjoy appreciations among specialised scholars.

[2] Actually, the Codex presented a numbering of the sheets, which was added with a stamp most likely at the end of the XIX century. While using it for convenience, Burns immediately clarified that "the numbering of the sheets unfortunately brings with it the psychological effect of attributing the present sequence a more definite meaning than it really has" (p. 22).

[3] The author refers to an overall revision of the ecclesiastical library heritage decided as a result of the Councils of Constance (1414-1418) and Basel (1431-1449). The revision included the drafting of new library inventories, and the restoration and binding of the most fragile materials.

[4] See Mappae clavicula. Alle origini dell’alchimia in Occidente. (Mappae clavicula. At the origins of alchemy in the West). Edited by Sandro Baroni, Giuseppe Pizzigoni and Paola Travaglio, Padua, Il Prato Publishers, 2013, pp. 46-47.