Sylvie Neven, The Strasbourg Manuscript. A Medieval Tradition of Artists' Recipe Collections (1400-1570)

Review by Giovanni Mazzaferro
Translation by Francesco Mazzaferro

CLICK HERE FOR ITALIAN VERSION

Sylvie Neven
The Strasbourg Manuscript
A Medieval Tradition of Artists’ Recipe Collections (1400-1570)
Londra, Archetype Publications, 2016

To define this book merely as the latest English critical edition (and the second integral one) of the Strasbourg Manuscript would be quite restrictive. The volume of Ms Sylvie Neven starts from the manuscript study (or, better said, of its copy) to reconstruct a set of other manuscripts (a "Tradition"; in fact, from now on the term will be used in this meaning with the capital "t") which shares with it part of their contents; in turn, the works that are part of the Tradition are contextualized within a genre which in German took the name of Fachliteratur (i.e. "specialist literature"). The Fachliteratur, covering medicine, alchemy, artistic techniques etc. is, fundamentally, a literature based on recipes, with its own syntax and features. The authoress studies the possible origins of this type of literature, the similarity between recipes of different nature, the areas of their origination, the possible authors and the drafters of the manuscripts; in workshops and in the few artefacts of that era which survived to the present day, she also searches comparisons with the findings of the techniques described in the texts. Sylvie Neven’s book is therefore, more than a critical edition, a fascinating journey into the world of medieval recipes.

The Strasbourg Manuscript

The Strasbourg Manuscript does not exist anymore. The specimen presenting him was preserved with the mark A VI 19 at the library of the Alsatian town, probably coming from the cards of the local St John's Commandery, founded in the fourteenth century (see p. 14). It was destroyed during the disastrous fire that hit the town in 1870 (purely incidentally, it should be remembered that it was not an unfortunate event, but one of the results of the Franco-Prussian War, the first episode of a series of conflicts for the control of Alsace and Lorraine, which only ended with World War II after the death of several millions of German and French citizens). Fortunately, in previous years, the manuscript had been studied by Charles Lock Eastlake, great expert and director of the National Gallery, and the author of the Materials for a History of Oil Painting in 1847. Eastlake published some excerpts in his collection dedicated to the history of oil painting. To better examine it, he commissioned a copy, which is kept at the National Gallery, in London, with mark 75,023 STR. We do not know anything of that copy, as we ignore who produced it, whether it was complete, and whether it complied with certain criteria required by Eastlake. Actually, Sylvie Neven states today, with reasonable probability, that it was a corrupt text, where some topics had been left out or otherwise moved from their original location. The first complete edition of the Strasbourg Manuscript was made by Ernst Berger, which made another copy from that of Eastlake and published it within the three volumes that displayed his 1897 Quellen und Technik der Fresko-, Öl- , und Tempera-Malerei. The first English complete edition, edited by the sisters Violet and Rosamund Borradaile, dates back to 1966. There were no other translations. 

The historical importance of the work was well understood since the time of Eastlake, who pointed out that, among the various recipes, there were some who also covered the production and the use of a drying oil for panel painting. The general structure of the manuscript was set by Berger with a subdivision into three parts, all independent of each other; the first consisted of a short series of recipes that the scribe said to have been transmitted by Henry of Lubeck; the second one of requirements dictated by Andrew of Colmar; in the third one the writer (we do not know who he was) spoke in the first person, and seemed to refer to his personal experience. The lost exemplary, according to palaeographic studies commissioned by Eastlake, dated back to the fifteenth century and was written in an ancient medium-Germanic dialect.

Soon - and not just for the issue of the drying-oil agent - the Strasburg manuscript became the prototype of artistic techniques proposed in Northern Europe around 1400; likewise Cennini’s Book of the art took a similar role with regards to Italy. While lost, the text has therefore become part of a very small group of reference manuscripts for the history of arti techniques, such as the texts of Heraclius, the De diversis artibus of Theophilus and the Mappae clavicula.

De-constructing for re-constructing

In order to explain what constitutes the novelty of the Neven issue, we are forced to open a parenthesis on the method. The authoress did not want to simply stick to the information acquired on the manuscript. First of all, she questioned the "paradigmatic" part of the text as a representative of all the techniques from Northern Europe. The reality is different, and speaks of hundreds of texts, each with its own specificity, depending in turn from different practices applied in different geographical areas. Therefore, she investigated and identified specificities one by one. Sylvie Neven belongs to a generation of historians of art techniques which has learned to analyse the particular always having in mind the big picture. One of these scholars (Mark Clarke) was responsible for the drafting, in 2001, of a formidable repertoire (The Art of All Colours. Mediaeval Recipe Books for Painters and Illuminators, also published by Archetype) where are scrutinized and classified over 400 medieval manuscripts written until 1500.

When the search field takes this size is inevitable to refer to electronic resources. Ms Neven is part of several projects for the registration of manuscript materials. I am quoting, among others, the Colour Context Website, under the auspices of the Max Planck Institute for the History of Science, in which more than 600 recipes were transcribed, up to the Theodore de Mayerne’s years (in short, until the beginning of 1600).

Source: https://arb.mpiwg-berlin.mpg.de/

As you can imagine, the critical point, when using these tools, is not so much to find data, but to scrutinise them according to uniform criteria. One is invariably faced with texts written in several languages, with materials and pigments indicated with different terminologies, with sometimes very high, and sometimes entirely superficial, analytical levels. All these texts, however - being part of that Fachliteratur which was mentioned at the beginning - have one thing in common: to be arranged through recipes, which are sometimes only one line, sometimes a whole page long. The fundamental task of those who work in the sector is to identify and separate every single recipe from each other. The normalization of a text entails some basic rules, the first of which is the separation of recipes related to each other. In fact, it often happens that a process of realization of a given material is immediately followed by the indication of a variant. In this case, the two recipes are to be separated. The scholar proceeds, in short, in a work of deconstruction of the text until the most elementary level is reached, and that level is precisely the recipe in the etymological sense of the word (from the Latin recipere – to take): “the recipe appears as ‘the shortest element in which the text could ultimately be decomposed’ (Halleaux, 1979: 74)… This definition could be refined by adding that the recipe is the smallest independent element into which the text could be divided. In fact, a recipe could be seen as an independent text in itself and could therefore be dissociated from its original recipe collection and introduced into the pages of another manuscript. For this reason, it may be argued that the format of the recipe could be considered as the structural unit common to several disciplines embedded within the manuscripts belonging to the Fachliteratur” (p. 55)

Once the de-composition has been implemented, it is of course to be checked whether other manuscripts provide sequences of "parallel recipes" with respect to the specimen covered (in this case the Strasbourg Manuscript) and then to determine whether there is a Tradition, constituted by more texts, which presents its contents. “Numerous recipes show similarities in their lexicon in describing a similar procedure. However, in order to be identified as part of the Strasbourg Tradition, a recipe would also need to be similar in terms of syntax. Syntactical similiarities would imply textual parallels that were sufficiently clear in order to attest the same textual origin(s). Therefore, parallel recipes probably derive from the same textual source and should be distinguished from recipes that are similar in their technical content (Clarke 2011: 17-18)" (p. 26).

The Strasbourg Tradition

Mind you, some scholars had already pointed out to the similarity between the Eastlake exemplary and other manuscripts, but this had not led to the systematic study of the Tradition. “The detection and defining of a textual and ‘technical’ tradition of recipe collection, as undertaken in the present study, could not have relied on serendipity – it resulted from the application of systematic research and analysis" (p. 27). “For the purpose of this study, only the recipe collections sharing at least 10 instructions in common were taken into account. From the initial large corpus of more than 600 examples, a smaller and clearer corpus of manuscripts belonging to the Strasbourg Tradition was defined. New textual evidence was discovered and the ‘Strasbourg Tradition’ now corresponds to a group of 15 manuscripts" (p. 29).

In sum, we must get used to thinking of two levels: on the one hand the single copy of the Strasbourg Manuscript produced by Eastlake and, second, 15 manuscripts that make up the Tradition; starting from here, we need to consider whether the examination of the Tradition may shed light on the single specimen.

The book examines which elements the manuscripts of the Tradition share. In general, it can be noted that the texts in question are all originating in southern Germany, and written mainly in three dialects: High Franconian, Bavarian and Alemannic (the latter is the case of the Strasbourg Manuscript in the sense strict). There are only few exceptions. We recall among them the Vossianus Chymicus Octavo 6 (p. 59), now preserved in Leiden, because we will mention it later. Mostly prepared in religious centres (typically in monasteries), all manuscripts appear as composite texts, in a twofold sense. First, they are bound together with other texts of Fachliteratur which are mainly concerned with provisions of a medical nature; secondly, the individual sections devoted to artistic recipes still seem to be a mix of recipes copied from other texts, of recipes provided for the occasion by local artists and of inputs collected (and sometimes even experienced) by those who wrote them.

The manuscripts, in short, share complete sequences of "parallel recipes". In certain situations some sequences may appear more complete than others. In particular, the 15 manuscripts of the Strasbourg tradition share five sequences of recipes; two of them are common to the Strasbourg Manuscript in the strict sense and the other documents. Sequence A has to do with the preparation of colours and materials for illumination; Sequence B provides guidance on how to temper pigments and to lay tints especially on wood.

Sequence A is demonstrated in the specimens that are considered to be (on a palaeographic basis) the oldest. However, it is fair to note that, if the absolute oldest appears to be the Manuscript Strasbourg (whose dating Sylvie Neven assesses to be before 1412, on the basis of information found in archives on Andrew of Colmar’s life - see p. 56 - ), the most complete sequence is presented in other manuscripts, especially the Vossianus Chymicus Octavo 6 and the Amberger Malerbuch (a transcript of the Vossianus Chimicus Octavo 6 is also provided in Appendix I). How is it possible that subsequent manuscripts would present more complete sequences? Manifold assumptions can hold. One thing appears certain: the Strasbourg Manuscript was not the prototype of Tradition; in fact, it displays textual shortcomings (indeed empty spaces, because the copied text was not interpretable) that can be completed through the use of the Vossianus. There was therefore an even earlier prototype. Having understood it, a lot can have happened: both the Strasbourg Manuscript and the Vossianus may be copies, relative to the sequence A, from the same prototype, and the editor of the Strasburg Manuscript may have decided to leaving out a part of it. However, it cannot be excluded that the process may have been much more recent and the waiver of part of the sequence may have occurred on the occasion of the copy which Eastlake ordered. It is to be noted, however, that, even in the part which was not discarded, the succession of the recipes in the Eastlake copy is different from that of all other traditional manuscripts, a clear sign that the copy was not faithful but coincided with an arbitrarily rearrangement of the available material. Finally, one should also note that the length of the Eastlake copy appears to be different from two descriptions of the burnt manuscript, which were discovered by Neven; with respect to them, the text appears shorter and arranged in a different way than the original.

Whatever may have happened, the Strasburg Manuscript clearly appears to be a transcript deriving in part from the copy of (at least) two previous texts: on the one hand a treatise on miniature, as precisely shown by Sequence A, which was later copied in many other witnesses of the Tradition and, on the other hand, a second treatise on painting (Sequence B) which instead appears to have had a much lower circulation and be shared with only two other manuscripts: one called Colmarer Kunstbuch and the Bamberger Malerbuch. Even in this case, the wider witness of the sequence is not the Strasburg manuscript, but the Colmarer Kunstbuch. Besides the two sequences in question, also recipes exist which appear sporadically in other manuscripts, as well as absolutely specific passages which, for this reason, are called by the Latin term unica.

The Mckell Medical Almanack in German, illuminated manuscript on parchment, Alsace, circa 1445, 12 leaves
Source: https://www.liveauctioneers.com

The new critical edition of the Strasburg Manuscript

Everything we have said before justifies how Neven came to the conclusion, on the basis of the texts of the Tradition, that the division of the Strasbourg Manuscript into three parts and ninety recipes, operated by Berger, should be reviewed. First of all the recipes (for the above described process of reduction to elementary units) become 114. Then, as part of the manuscript, the author identifies seven different sections. However, it should be immediately clarified: no shift, for whatsoever reason, in the order of presentation of the recipes was operated. Rightly so, Ms Neven limits herself to report - in her opinion - the original order of the prototype, which was then historically modified, for many reasons.

Let us examine all the seven sections (see pp. 39-44):

• Section I: corresponds to what, according to Berger, was the first (and short) part of the manuscript. It presents a short treatise attributed to Henry of Lubeck with guidance on techniques to paint in miniature;
• Section II: is the Sequence A which was discussed earlier. In the Strasburg manuscript, Sequence A is interrupted by two recipes that do not appear in any other text.
• Section III: it consists of only recipes 16 and 17, which are preparations for gilding. Between the two recipes the sentence is displayed: "This Andres von Colmar taught me." Berger closed the first part of the manuscript with the recipe 16 and started the second with the recipe 17. He attributed to Andrew of Colmar the paternity of twenty recipes, then moving on to the third party. According to Neven, things were different. In the Eastlake copy, recipes are proposed one after another without any spacing. The hypothesis is that the indication of Andrea of Colmar’s name was actually a note in the margin of the burned manuscript, placed between the recipes 16 and 17, and it was referred only to these two; in fact, thereafter, the text continues with prescriptions belonging to the Sequence A. The explanation by Neven seems absolutely plausible.
• Section IV: in substance, it is the least consistent of all, because it presents varied recipes that begin with an indication of the scribe that he would start providing instructions on painting techniques practiced in Lombardy (Eastlake thought it was not Lombardy, but London). Only a few of these recipes - which concern binders, translucent colours, preparation of the scrolls, gilding and various types of paint - also appear in other manuscripts of the Tradition and without being tied in sequence. The author supposes that these were contributions of the scribe. Within Section IV (which is then actually divided into two subsections IV a and IV b) appear the recipes of Sequence B. Allow me to still report an aspect which, in my opinion, remains unresolved: why Lombard techniques (or London ones, according to Eastlake)? We can assume that the anonymous scribe could have travelled there (or be born in those places).
• Section V: presents the recipes of Sequence B, common to the Colmarer Kunstbuch and the Bamberger Malerbuch.
• Section VI: it shows information relating to the preparation of supports not used in painting and illumination (for example, metal plating).
• Section VII: the last section includes recipes that appear to come from other sources. For instance, the names of Theophilus and Pietro di Sant'Audemaro have been suggested. In fact, the unique characteristic of this part is its heterogeneity, so that it includes also instructions for the preparation of soaps and a couple of magical prescriptions. 

A literary text or a workshop recipe book?

The final chapter of Neven’s work is devoted to an examination of what, ultimately, is the main question every time one has to do with a medieval recipe book. Are we faced with a text that was copied to pass a literary tradition and therefore was not used in the workshops of craftsmen of the time? Or is it a collection of a purely technical nature, arising directly from the knowledge of the artisan, and aiming at the education of pupils? The topic has been discussed for more than one century. Neven proposes theses for and against both alternatives and, ultimately, already at p. 4, provides the answer referring to Mark Clarke’s teachings: since the two types, in substance, coexisted, we need to manage distinguishing today the literary and the teaching texts. The analysis has to be all over the field, starting right from the appearance of manuscripts, and when it came down to us: some manuscripts, such as the Bamberger Malerbuch, appear written in an orderly and refined form, embellished with titles and headings written in red. It is difficult to assume that these texts were intended to be used near a furnace, and it is simpler to think that they were intended for a scholar context. On the other hand the Trierer Malerbuch (we are talking about another manuscript belonging to the Tradition of Strasbourg) looks much more unkempt and messy, with the interpolation of subsequently written recipes among the spaces left empty (see pp. 63-68).

The same could be said about the scribes. Not always (in fact, hardly ever) one is able to figure out who made the copy; many inconsistencies in the texts could hint to the technical unpreparedness of those who drafted the copy (in the end, it is what could have happened with Cennino Cennino’s Libro dell'arte, whose oldest copies reaching us was carried out in the Stinche prison in Florence in 1437). And, nevertheless, one could not explain the recipes - which do exist - in which writer's contributions are presented. Nor can it be ignored that in some situations we know the name of the extensor. It is the case, for example, of the Liber illuministarum (yet another witness of the Strasbourg Tradition), extended by Konrad Sartori, known not only for being the librarian of the monastery of Tegernsee (Bavaria), but also a copyist and illuminator (see. p. 63).

The specificity of the Strasburg Manuscript

In this delicate game between literature and technique, it is of course of particularly importance the analysis of the specificity of a text. These particularities appear to be linked to the territory in which the text itself was produced. In the case of the Strasbourg Manuscript, then, it was the area roughly comprising Alsace and Bavaria (once again, the authoress denies the possibility that the manuscript can be taken on the prototype of all Northern European techniques). A peculiarity is without doubt constituted by the oil drying agent for panel painting, already highlighted by Eastlake in 1847. Neven identifies a second specificity in the use of anthocyanins, or water-soluble pigments of plant origin from poppies, cornflowers and blueberry plants. Since they are particularly sensitive to the Ph, it was believed that, for this reason, anthocyanins were used exclusively for the clothing and textile colours, while the Strasburg Manuscript certifies their use in painting and miniature. The last pages of the work - as well as Appendix II - are dedicated to laboratory experiments that aim at studying the production of anthocyanin pigments, by following the recipes provided in the manuscript, and find traces in artefacts of the first 1400s in the Strasbourg area, which have come down to us.

A further proof, if ever there was a doubt, of how wide the spectrum of knowledge is of which historians of artistic techniques must be equipped with, ranging from palaeography to chemistry, from history to the knowledge of ancient dialects and philology; this would almost induce us to say that a good historian of artistic techniques is not a specialist, but rather a new (perhaps the last) humanist.