FORGOTTEN MARGINALIA AND THE FRENCH AND LATIN MANUSCRIPT TRADITION OF LE MIROUER DES SIMPLES AMES BY MARGUERITE PORETE MARGINALIA OLVIDADOS Y TRADICIÓN MANUSCRITA FRANCESA Y LATINA EN LE MIROUER DES SIMPLES AMES DE MARGUERITE PORETE

This article advocates a fresh critical study of the manuscripts of the “heretical” book, Le Mirouer des simples ames by Marguerite dicta Porete, and an examination of codicological evidence which neither the standard editions nor the modern translations take into account. It argues for analysis of the codex traditionally known as the Chantilly manuscript (Musee Conde, ms. F xiv 26, cat. 157) and of the manuscripts of the Latin branch preserved in the Vatican Library (Biblioteca Apostolica Vaticana, Vat. Lat. 4355; Cod. Rossianus 4; Cod. Chigianus B IV 41; Cod. Chigianus C IV 85 and Vat. Lat. 4953). The analysis focuses on the marginal marks ( maniculae, nota bene and iconography) which demonstrate that those books were used by active readers. The article also highlights the way in which the marginal marks provide us with information about the transmission and reception of the Mirror , and classifi es them according to function, depending on whether they facilitate access to, evaluate or even contribute to the meaning of the text of the Mirror itself.


INTRODUCTION
By raising the question of annotation, we place the study of medieval literature in its medieval context as a cultural artifact 2 .
As we know, the 2010 article by Robert Lerner, New Light on The Mirror of Simple Souls, re-opened scholarly debate concerning the reliability of the different versions of Marguerite's supposedly heretical book 3 . What seems clear following Dr. Lerner's observations about some suspicious variants in the only complete French version that we have (traditionally known as the Chantilly manuscript) is that we must now 2 S. Nichols,Sociology of Medieval,p. 47. 3 R.E. Lerner,New Light,  manuscript (Musée Condé,ms. F xiv 26,cat. 157) and of the manuscripts of the Latin branch preserved in the Vatican Library (Biblioteca Apostolica Vaticana,Vat. Lat. 4355;Cod. Rossianus 4;Cod. Chigianus B IV 41;Cod. Chigianus C IV 85 and Vat. Lat. 4953). The analysis focuses on the marginal marks (maniculae, nota bene and iconography) which demonstrate that those books were used by active readers. The article also highlights the way in which the marginal marks provide us with information about the transmission and reception of the Mirror, and classifi es them according to function, depending on whether they facilitate access to, evaluate or even contribute to the meaning of the text of the Mirror itself. return to the origins of the Mirror and focus our efforts both on how the manuscripts developed in relation to each other, and on the materiality of the codices 4 . We align ourselves with this philological revival: we have already published the results of our research on the materiality of Ch 5 and in the next pages we will move forward to compare the manuscripts in the French and Latin traditions through a close study of their margins. Essentially, we intend to highlight the importance of taking a new, deeper look into every copy of the Mirror as a singular physical object with its own unique context and set of uses.

Keywords
In doing so, we will consider six of the main manuscripts in the Mirror tradition, paying closer attention to their "virgin" marginalia 6 . We will collate the information that we have about Ch, comparing it with the new material that we have collected from the margins of the documents of the Latin tradition, which are preserved in the Vatican Library. In other words, we will compare the main French codex with fi ve of the six manuscripts on which Paul Verdeyen based the standard edition of the Latin text, referring to these following his labels: A, B, C, D and F (A = Biblioteca Apostolica Vaticana, Vat. Lat. 4355; B = Biblioteca Apostolica Vaticana, Cod. Rossianus 4; C = Biblioteca Apostolica Vaticana, Cod. Chigianus B IV 41; D = Biblioteca Apostolica Vaticana, Cod. Chigianus C IV 85; F = Biblioteca Apostolica Vaticana, Vat. Lat. 4953) 7 . We will attempt to understand them as cultural artifacts that were used in a specifi c historical context. This is a standpoint which forces us to take certain considerations into account. We will argue that the different Mirror versions can be understood not just as linguistic and geographic variants, as the translations and copies of an unknown prototype, but as specifi c reading devices that were viewed and handled by their historical users in specifi c ways 8 . 4 Lerner discusses the reliability of Ch, comparing it and the English version with some fragments in Old French discovered in the Valenciennes Public Library by G. Hasenohr Let us start our analysis by stating that since Guarnieri and Verdeyen's editions of the Mirror we know that at least Ch, A and D all possess marginalia 9 . Indeed, when Guarnieri described Ch in her preface to the 1965 edition, she acknowledged that: di varia natura sono le numerose postille marginali (the numerous marginal notes are of different kinds) 10 and this was followed by a list of examples of the various "kinds" or "functions of the notes". In the same way, on the next page of the same edition, Verdeyen writes about the A codex, the main Latin manuscript [our translation]: From the beginning to the end, we fi nd marginal notes which summarize the content or highlight some locutions of the text. Our edition has not taken into account those reader notes 11 . Guarnieri and Verdeyen's simple act of indicating the existence of marginal material suggests that both editors were aware that by omitting to publish the notes they were losing certain information.
In fact, Verdeyen demonstrates the main problem, which is that from a traditional philological point of view, the information contained within the margins of medieval manuscripts has often been treated as if it were of lesser status than the information framed between the margins 12 . In addition, the marginal markings tend to be described as mere reader marks. In this sense, editing a medieval book such as the Mirror has frequently been a process of cleaning up the text and, as a result, dissociating it from the immediate historical contexts in which scribes wrote, revised and annotated specifi c manuscripts, that is to say, from its manuscript culture 13 . This is a prejudice which has been corrected in recent decades by scholars working in the New Philology or the History of Reading schools of thought 14 and it is one that we continue to examine in relation to Marguerite Porete's manuscript tradition in particular.

MARGUERITE'S MANUSCRIPTS IN A DEVOTIONAL CONTEXT
Some months ago I went to Rome on a research trip to the Vatican Library, where I set out to fi nish my transcription of the marginalia in manuscript A, which Verdeyen mentions. Afterwards, I undertook further work with the other Poretean codices preserved in the Library and was able to confi rm that every single manuscript in the Latin branch possesses marginalia, but each to a different degree. After several days of study I was convinced of the need to consider all the manuscripts afresh. Two essential questions seemed to arise: a. Why do these marginalia exist? That is to say, is their function important for the existence or working of the Mirror?
b. Can we somehow classify these manuscripts from the point of view of annotation?
Let us start with the second question: we can assert that the format of these books is as important as their date, since the combination of both features provides us with a good overall idea of the historical nature of the codices. Let us take a look at the following diagram 15 : We can observe here what Guarnieri states in her 1986 codicological description: the Chantilly manuscript originated somewhere in the Orléans region in the late 15th century 16 . On the other hand, all of the Latin manuscripts came from Italy: three of them (A, B and C) were circulating in the 14th century, one (D) in the 16th century and the last (F) in the 15th century 17 . The majority of the books are small compared to the only one that does not seem to follow the standard format, manuscript F, which is considerably larger than the others. As Justine Trombley describes it, this manuscript 18 is made up of documents relating to the negotiations with the Greek Church at the Council of Florence (1438-1439)[;] on folios 29r-32r there is a list of thirty direct quotes taken from a Latin Mirror copy which are presented as errors and are followed by refutations which use Scripture and Theological/Legal authorities to point out precisely why the extracts are erroneous.
As Trombley concludes, we don't know exactly how the documents found in manuscript F were used, but it is clear that there was a persecutory context in which the Mirror was considered a source of doctrinal errors 19 . So this book was not designed for devotion, but for controversy. If we add to this the miscellaneous character of F, the glosses which follow every fragment of the Mirror text and, above all, its high degree of legibility (we will return to this point) we can explain the almost total absence of marginal notes.
So, except for F, we have here a group of fi ve little manuscripts from the Low Middle Ages. Their size, the evidently cheap materials with which they were made and the lack of sophisticated illumination suggests, as Justine Trombley asserts, that [t]hey were clearly made for practical use 20 . This calls to mind how important it is to understand these books as belonging to a period of time in which a series of developments in reading tools permitted readers to read in silence, making use of just their eyes and their intellective faculties 21 . We are in front of precious objects which allowed medieval readers to penetrate Marguerite's text individually.

THE MIRAGE OF THE "READER MARKS"
Having located these books in their devotional context we can now determine what the marginalia represent. As we have said, the main problem of medieval marginalia is the tendency to label them as "reader marks", and in so doing, neutralize all attempts to understand the complex nature of these signs. In fact, the homogeneity of the term is problematic, because it hides a variety of functions, correlations and implications that form part of the praxis of premodern reading which we will try to explain in the case of these fi ve manuscripts.
In our research we identifi ed two sets of interrelated functions. In order to illustrate these clearly it will be useful to separate the fi ve codices into two groups: If we accept that both groups represent devotional tools giving access to the same work and that both were produced between the 14th and 16th centuries (a rich period of time in terms of developments of medieval reading practices), then we have to ask ourselves why the margins in the second group are so much cleaner. The answer is related to devotional reading habits during the last centuries of the Middle Ages. The two manuscripts in the second group (B and C) are better prepared for silent reading (and in this sense they seem more expensively produced than the others): they are decorated with capital letters, red and blue fl ourishes, chapter title divisions, pilcrow signs, space between paragraphs, and other aids for the reader 23 . We might describe them as "clean" because they are "clear" to the reader. On the other hand the lack of visual reading aids has encouraged more extensive marginal annotation in the fi rst group.
Since it was fi rst published, readers have said that the Mirror is a diffi cult work: doctrinally it is speculatissimus 24 , as we read in one of the codices; at least in a silent reading, from the beginning to the end, the structure is more thematic (or rhizomatic) than rational; and, in the codices of the fi rst group, the copy is cheaply produced and reading aids are scarce 25 . We fi nd annotations that try to make up for these barriers to legibility: visual interfaces that attempt to facilitate access to the text. Here we classify this kind of marginalia as "instrumental", with some examples from the codices of the fi rst group:  As we have said before, the most important function of this kind of annotation is to make the text readable. Zan Kocher shows in his most recent book that we cannot understand the Mirror without comprehending the implications of the allegory as a means of producing sense 26 : we could say the same applied to the medieval readers of those manuscripts, and a high percentage of A and D annotations demonstrate this. Let's look at an example in A at the beginning of Chapter 25th, f. These two notes attempt to restore the original use of dialogue in conveying allegory, and in this sense, they articulate the Mirror for the reader's benefi t, solving the problem of the visual comprehensibility of the manuscript and, at the same time, marking Ratio's words as a question. However, while restoring allegorical dialogue is the most important purpose of the functional notes, as we can observe above, there are more:

a.
Locating an element in a list b.
Locating an image or a mnemotechnic element c.
Locating a topic.
For a and b let's consider ff. 30v-31r, which contain part of the 30th and 31st Chapters: the text-image and go on to list each one of the sections: on the one hand, this enables easier access to the text and, on the other, it reinforces its mnemonic structure. In Ch we will see an example in which the parts are not merely listed, but explicity drawn 29 .
Concerning point c there are numerous examples of marginalia used to draw attention to a topic, perhaps due to the lack of linearity in the Mirror's "plot", which makes it very diffi cult to locate any specifi c passage otherwise. The manuscripts are full of these notes. For example, in A on f. 22v, we fi nd: This note comes at the beginning of the 59th Chapter, which develops the topic of the three deaths. It fulfi ls two different but closely related functions: fi rst, the annotator uses it to mark a specifi c passage (that is, to locate an element); at the same time, they challenge the reader to pay particular attention to this point, thus making a positive evaluation of the text (since it is impossible, in fact, to mark the text without evaluating it at the same time). This second purpose would fall under the branch of hermeneutic functions.
As we have seen, part of this branch of functions is related to what we have traditionally called nota bene. We have to remember that a nota bene is not just a location mark: it constitutes a reading guide that informs the text for both the annotator and future readers 30 . This is clear in notes where the annotated topic or the fragment is explicitly evaluated using an adjective, an adverb (like "p[u]lcre" in one of the preceding examples) or, as is the case at the top of f. 15r, an entire expression 31 :

= O qua[m] b[e]n[e] d[ici]t[!]
Since this kind of evaluating note could be used as a reading guide, it suggests specifi c instances of readers accessing the text and it also builds the text, which is to say that the interface in the margins provides a framework for reading the Mirror in a defi nite way. This capacity of the marginalia to transform the main corpus of the Poretean book could be exemplifi ed on folio 22v, as we saw above, in the lower part of its margin. Concerning the "triplici vita" (or the "three deaths") we read: What we have here is the explicit transformation of a prose text into schematic form. The schema is based on the segmentation of the mystic way proposed in different parts of the book. As we know, the experience depicted in the Mirror is strongly structured: these kinds of reconstruction attempt to keep this inner organization clear and, at the same time, create a stronger mnemonic text through parallelism, rhythm and repetition.
Following these kind of reading practices in devotional treatises, in the incunabula period such interfaces were incorporated into the printed book as a part of its reading device 32 : these marginal reading interfaces, printed and manuscript, are pretty much alike and both work in the same way from the reader's point of view. The main problem arising in relation to hermeneutic annotations in Ch, A and D is whether we should interpret them as "mere" reader marks or as an integral part of the manuscript copy. In our current phase of research we are not yet able to confi rm whether one possibility is preferable to the other, but it does seem clear that a wider look at the manuscript provides relevant information.
The case of Ch allows us to observe that the nota bene are distributed from the beginning to the end of the codex, but not with uniform frequency. The fi rst annotation appears on folio 9v and the last one on 117r, but they are clearly concentrated in the fi rst half of the manuscript (up to folio 72r). This uneven spread can be explained following the working hypothesis whose feasibility we are in the process of testing right here. Fig. 8. Chantilly manuscript [Ch], ff. 29v and 30r. Photo: Zan Kocher, 1994. This hypothesis is concerned with the extreme homogeneity which Ch shows as a manuscript. It suggests that its margins might contain annotations that also existed in a previous French version: in fact, paleographic analyses of Ch, A and D have shown that the date of writing of most of the notes and the main corpus coincides 33 . This group of annotations would have followed a copy-invasive process described by W. Schipper 34 : These [kinds of] annotations must at one time have been incidental in the sense that they were simply added in the margins because a reader wanted to mark the passage for himself, but they were copied along with the main text into new copies of the book, and in the process became a part of the book itself, instead of remaining strictly marginal and peripheral.
The original group would have been augmented with the private annotations of the subsequent Ch readers, who used different private marks (in this case, customized maniculae or nota bene diverging in format) to indicate the passages which interested them in particular. This invasive process of 33  adding private notes would have been incorporated into later copies as a part of the reading device of the book 35 .
In fact, from a synchronic perspective, we can talk about one reading device, because what makes it extremely diffi cult for a modern researcher (as it did for an earlier reader) to distinguish between the different reading strata in Ch is the strong tendency towards uniformity which the codex presents. This homogeneity is present both in the appearance of the book as an object, in large part thanks to the distinctive interface or group of interfaces that it contains, and in the reading strategies that the use of one or combined marks in the book implies.
Two possible explanations for this homogeneity arise: fi rst, these signs could have been added in a relatively short period of time, within a community that employed similar reading strategies 36 . Secondly, such uniformity might have been intentional on the part of copyists who wished to emulate the style and tools of some early printed devotional books. Since the production of Ch coincides with the incunable period this second possibility is very interesting.
In conclusion, two processes appear to have affected Ch, A and D: an invasive process and a homogenizing one. This in turn produced the uniform look that characterizes these manuscripts, mutating them into more complex objects with different reading estrata and highly visible markers enabling access to the text.

CONCLUSIONS
By way of conclusion, we present here some short refl ections on the materials we have been examining. 4.1. Further research will be required concerning the "instrumental notes", focusing on elements of the Mirror manuscripts that are not described here and looking at them as objects that existed in a specifi c historical context. For this research, it will be necessary to evaluate every codex in its particularity: this is the only way to understand the different Mirrors in their real contexts.
4.2. Concerning the "hermeneutic annotations" as reading guides, we now have enough material to write a new chapter on Marguerite Porete's reception and in fact it seems essential that we do so. In a case like 35 The idea of annotation as an "invasive process" is developed in R. Hanna III,Annotation as a Social,p. 182. 36 This idea appears to fi nd support in a nota bene in Ch, f. 108v, in which we can read: "Notez bien bonnes pucelles". Cf. my analysis in Los marginalia, p. 267.
Marguerite's, in which the documentation is so scarce, we cannot discount such precious information just because it is anonymous or marginal. As reading devices the nota bene can show us how the historical readers used to access the text and simultaneously how they evaluated certain parts of it. If, as seems to be the case, the annotation also contributes to a specifi c reading of the Mirror, then devotional reading, which we can describe as guided reading, seems to coincide with the mystagogy. 4.3. Finally, it is surprising that the majority of notes evaluate the text in positive terms (it is always essential to remember that, from the point of view of the annotation, the readers seem to have considered them devotional and not heretical books) and that the negative ones are amazingly coincidental across the different copies. This would allow us to answer several questions: what were the topics in which the readers were interested? How were certain passages, which today seem "dangerous" or "heterodox" to us, read in a strictly medieval context? How was the reader guided to read those passages through the notes? Did the readers understand Marguerite's humour? And the questions continue.
Only by reading the Mirror historically will we come closer to Marguerite's words.