The Oldest Manuscript Tradition of the Etymologiae (eighty years after A. E. Anspach) (pdf)
(*En Portugués aquí; En Español aquí; En français ici)
The Etymologiae of Isidore of Seville was one of the most widely read works of the early Middle Ages, as is evidenced by the number of surviving manuscripts. August Eduard Anspach’s handlist from the 1940s puts their number at almost 1,200, of which approximately 300 were estimated to have been copied before the year 1000. This article, based on a new manuscript survey of the early medieval manuscripts transmitting the Etymologiae, brings the number of known surviving pre-1000 manuscripts transmitting the Etymologiae to almost 450. Of these, 84 well-preserved codices and 24 fragments contain the canonical Etymologiae, i.e., they reflect the integral transmission of Isidore’s work as an encyclopedia, while 300 well-preserved codices and 21 fragments reflect the selective or non-canonical transmission of the Etymologiae, principally not as an encyclopedia. Due to the uneven survival rates of manuscripts of canonical and non-canonical Etymologiae, it seems likely that the latter accounted for perhaps as much as 80-90% of manuscripts transmitting Isidore’s work before the year 1000. Four non-canonical formats emerge as having been particularly influential in the early Middle Ages: the separate transmission of the first book of the Etymologiae as an ars grammatica; the compilation of various catechetical collections, sometimes in question-and-answer form, from books VI, VII, and VIII of the Etymologiae; the incorporation of material from books V and IX into law collections; and the incorporation of segments from books III, V, VI, and XIII into computistic manuals. The surviving manuscripts suggest that the latter format emerged in the insular world, while the others are more distinctly Carolingian. Northern France and northern Italy emerge as the two most important regional hubs of the copying of the Etymologiae in the ninth and tenth centuries. While in the former region, non-canonical formats seem to have been the most important vehicle of the transmission of material from Isidore’s work, in the latter, the canonical format may have been more influential, indicating that there existed regional differences in the reception of the Etymologiae.
It is difficult to begin any treatment of the medieval manuscript tradition of Isidore of Seville’s oeuvre majeure otherwise than with what may sound like a eulogy. The influence and popularity of the Etymologiae in the European Middle Ages has been emphasized innumerable times, and so it can only be reiterated here that this work was among the most important medieval works and at times was copied so widely that every intellectual center either possessed a copy (or two, or three) or at least, in a desire to own one, obtained a selection from it. The prodigious number of surviving medieval manuscripts containing the entire 20-book corpus or its parts is alone an undeniable testimony of the tremendous significance of the Etymologiae in the European Middle Ages. Yet, the fact that more than a thousand manuscripts transmitting Isidore’s most important work survive from the seventh to fifteenth centuries also proves to be a substantial barrier to the full understanding of its manuscript tradition. It is one reason why we still lack a coherent and detailed critical edition, much less a satisfactory assessment of the transmission, reception, and appropriation of the Etymologiae in the Middle Ages. As the continued work of Jacques Elfassi shows, we have not yet even accounted for all of Isidore’s sources.
How daunting the task of editing the Etymologiae is and that it cannot be achieved by the effort of any individual scholar has been demonstrated by the example of August Eduard Anspach (1860 – 1943). In 1912, Anspach, a professor of Classical Philology at the University of Freiburg im Breisgau, was tasked by the Corpus Scriptorum Ecclesiasticorum Latinorum in Vienna to produce a scholarly critical edition of the Etymologiae that would supplant the recently appeared “school” critical edition of Wallace M. Lindsay. Well-trained philologist that he was, Anspach began by traveling through European libraries, seeking out manuscripts, assessing their relative value for edition-making, and collating them. While Anspach produced several valuable studies in the subsequent thirty-one years, neither the edition nor any substantial study of the manuscript tradition of the Etymologiae saw light of day before his death in 1943. All that remained of Anspach’s work were extensive handwritten notes about the manuscripts of the Etymologiae and several other Isidorian and Visigothic works. In the 1960s, it was agreed that Anspach’s estate would be moved to the newly established Centro de estudios e investigación San Isidoro in León. In 1966, José María Fernández Catón, the director of the diocesan archives of León, produced two publications from Anspach’s legacy: a catalogue of his estate and, more importantly, a handlist of the manuscripts of the Etymologiae.
Eighty years after Anspach’s death, Fernández Catón’s Las Etimologías en la tradición manuscrita medieval estudiada por el Prof. Dr. Anspach remains the most extensive resource for the study of the manuscript tradition of the Etymologiae. In what follows, I use it as a point of departure for a survey of more than 400 pre-eleventh-century direct witnesses of Isidore’s most important work and provide some observations on the early medieval manuscripts of the Etymologiae. Because of the large number of manuscripts concerned, the perspective from which this article is written is that of an eagle rather than that of a mouse. We shall observe a large corpus of material from a distance, scanning it for general patterns and paying attention to the most outstanding trends rather than approaching the over 400 manuscripts at close proximity to examine each of them in great detail. In other words, this article focuses on what is general and characteristic, rather than what is unique and particular, hoping to paint the broad contours of the pre-1000 history of one of the most important medieval texts.
The early medieval manuscripts of the Etymologiae in Anspach’s/Catón’s handlist
Let us begin with the basic numbers. Anspach’s/Catón’s chronologically organized handlist contains 353 manuscript items (under 293 separate catalogue numbers) assigned chronologically from the eighth to the turn of the eleventh centuries. Baudouin van der Abeele, who has produced the most substantial summary of this handlist, puts the total number of items at 1,080, which means that Anspach believed almost a third of all manuscripts that contain the Etymologiae to be early medieval. In Anspach’s reckoning, 89 of these manuscripts transmitted what may be termed the “canonical Etymologiae,” that is, the complete 20-book encyclopedia or its substantial part (e.g., only the first or the second half of the work). The remaining 264 items in Anspach’s/Catón’s handlist are classified as “fragments,” that is, as transmitting a selection from Isidore’s work (see Tab. 1).
Tab. 1: Overview of the early medieval witnesses of the Etymologiae according to Anspach’s handlist.
While Anspach’s/Catón’s handlist provides a useful benchmark for quantifying manuscripts of the Etymologiae in the absence of better overviews, it also suffers from many problems. First, quite a few items on Anspach’s list are duplicated or even triplicated. Anspach also misdated many manuscripts, so that he both included post-1000 manuscripts among early medieval ones and excluded some pre-1000 manuscripts by incorrectly dating them to more recent centuries. It is clear that he was not always able to examine manuscripts in person, but rather relied on manuscript catalogues or reports from his peers, and as a result, included many items that were identified incorrectly as containing the Etymologiae. In yet other cases, he decided to include works that quote or reuse the Etymologiae, such as the Liber glossarum. One manuscript included in Anspach’s handlist can no longer be identified using the information provided. Eight items were destroyed in the course of the twentieth century. Last but not least, a substantial number of early medieval manuscripts are missing from Anspach’s handlist because they contain anonymously transmitted excerpts, material notoriously difficult to identify even today, much less in Anspach’s days.
Thankfully, the last eighty years have seen the appearance of many research tools that allow one to identify manuscripts that Anspach could not, in particular the Codices Latini antiquiores for pre-800 manuscripts and Bischoff’s Katalog der festländischen Handschriften for ninth-century material. As a result, Anspach’s numbers can be substantially amended, although the search for excerpts in miscellanies, florilegia, handbooks and similar knowledge collections remains open-ended, as does the hunt for fragments, which continue to be unearthed in libraries and resurface in private collections. For this reason, the overview provided below is necessarily incomplete, principally on the side of what Anspach would have called “fragments.” It should be considered an improvement of our current knowledge but certainly not the final word on the early medieval tradition of the Etymologiae.
A re-evaluation of Anspach’s/Catón’s handlist in the light of newly available resources has yielded 443 manuscripts fully or partially containing the Etymologiae that date from the seventh to the early eleventh century. This number depends on the application of several criteria. In the first place, I excluded manuscripts of works based on, citing, or reusing the Etymologiae which are transmitted in their own right (as is the case with the Liber glossarum, Hrabanus Maurus’s De universo and several grammatical works from the Carolingian period borrowing from the first book of the Etymologiae). I also excluded manuscripts in which the material taken from the Etymologiae is significantly altered or reworked, so that, while Isidore’s knowledge corpus can be recognized as its source, they cannot be seen as direct witnesses of the text of Isidore’s work (e.g., glossaries and word lists dependent on the Etymologiae). One of my criteria was that a manuscript must contain a minimum amount of material to be included, my limit being the presence of at least one section of a chapter from the Etymologiae as defined by Lindsay’s edition.
As with all corpora, ambiguity is present. In several instances, for example, chapters or entire book sections of the Etymologiae were attached to other works as paratexts in the Carolingian period. How should we treat these witnesses of the Etymologiae, especially since building a manuscript corpus implies including all manuscripts containing the same paratext, even if they are copies descending from the same manuscript? Should anonymous epitomes or excerpt collections built from bits and pieces of the Etymologiae that bear a new title be treated as new works, just as the Liber glossarum, even when they appear in fewer manuscripts and their autonomy is less clear-cut? What should we make of small snippets of the Etymologiae copied on flyleaves, squeezed into blank spaces, inserted on slips in between folia, or copied in the margins of other texts? These instances of the appropriation of Isidore’s most important work have been included, but they could just as easily be excluded by someone else.
The Big Isidore and the small Isidores
When Anspach distinguished full manuscripts of the Etymologiae from “fragments,” it was to recognize an important trait of the manuscript tradition of this work, namely, that it circulated both as a whole, in agreement with the design of the Sevillian bishop, and as various selections of this whole, chiefly in response to the needs of its medieval users. Naturally, many if not most medieval works were excerpted at one time or another, but in the case of the Etymologiae (and certain other texts), the phenomenon reached such proportions that it had a transformative effect on the text itself, at least in the early Middle Ages, when it may have been at its peak. This is not just a question of excerption; diverse strategies of repurposing and appropriation often resulted in the emergence of new textual entities, including those that lived for centuries as Isidore’s secondary works alongside the canonical Etymologiae and defined how Isidore would have been read and interpreted. Due to their perceived derivative nature, these entities have rarely elicited scholarly interest, even though they inform us about the tastes and perspectives of medieval users. Since the study of the manuscripts of the Etymologiae has been mostly subordinate to the task of editing and given the large number of surviving manuscripts, it is understandable that the manuscripts of the complete text were privileged. However, it is crucial to stress that the guise in which modern scholarship knows the Etymologiae thanks to its editors, that is, as an encyclopedia (moreover, in a Teubnerian pocket format), was not the form in which this knowledge corpus was accessible to the majority of its medieval readers. Many early medieval users would not think of Isidore as an encyclopedist at all, but rather would consider him to have been a grammarian, a computist, a medicus, or a zealous bishop keen to reform clergy and promote law, depending on what kind of manuscripts they handled and whether they progressed far with their training in letters and Christian doctrine.
To be clear, the canonical Etymologiae is not merely a construct of modern editors. It represents an important strain of manuscript tradition descending directly from that Etymologiarum codex nimiae magnitudinis mentioned as left behind by Isidore in the Renotatio of Braulio of Zaragoza that follows the 20-book division imposed on this text by the same Braulio. As such, it represents a particular venerable embodiment of the Etymologiae which was recognized as such by medieval copyists as well as readers. This embodiment was typified not only by the content, order, and division of text, the presence of particular paratexts (the letters exchanged between Braulio and Isidore as a preface, a general list of libri at the beginning of the text, and lists of capitula attached to individual books or book sections), but in a period in which texts existed strictly as material objects, it was also characterized by a particular physical format. The early medieval manuscripts of the canonical Etymologiae are usually large-format and relatively bulky codices. They come either in one or two volumes (even though the handy division into 20 books allowed, in theory, for different modes of segmentation), and represent a specific kind of a book – the high-end library codex which is virtually the only form of the Etymologiae to sport decorated initials and other forms of ornamentation.
By virtue of possessing these traits, codices of this Big Isidore represent a discernible homogeneous group among the early medieval manuscripts of the Etymologiae that stands in contrast to the many small Isidores, manuscripts transmitting the Etymologiae selectively. Importantly, a Big Isidore is not defined only by its content, structure, and physical properties, but also by a particular kind of intended readership and context of use. The extant body of scholarship on the manuscripts of the Etymologiae has neglected the questions of who these Big Isidores were produced for, who had access to them, how they were used, and what concerns governed the decision-making of copyists or scriptoria when obtaining or producing a Big Isidore or opting for a small Isidore instead. It is, for example, noteworthy that early medieval Big Isidores seem to have been designed as in-house books produced for the internal use of particular communities. The books remained remarkably static, so that even today, as long as a given manuscript collection was not dispersed in the last few centuries, many early medieval Big Isidores are to be found in the same places where they were produced.
Unlike the Big Isidores, the small Isidores do not represent a homogeneous group of manuscripts or an entity that would be recognized in the Middle Ages. Rather, it is a catch-all category useful to contrast manuscripts transmitting the material from the Etymologiae selectively with those containing the canonical Etymologiae. Nevertheless, while this heterogeneous category features notable exceptions, small Isidores display certain uniform tendencies. Although there are examples of high-end library books among them, most small Isidores were produced in modest circumstances and share a particular down-to-earth quality. On average, they tend to be smaller in size than Big Isidores and include some pocket-sized specimens. Indeed, one property that distinguishes small Isidores from Big Isidores is their portability which contrasts with the static nature of the Big Isidores prescribed by the size of the canonical Etymologiae and the technological constraints of manuscript production in the early Middle Ages. Selection was one of the ways of giving Isidore’s extremely useful knowledge corpus wings. Furthermore, while the small Isidores are to an extent an artificial category, sifting through them reveals several coherent clusters of manuscripts that represent established formats of the non-canonical Etymologiae and, therefore, resemble the Big Isidores in their homogeneity of transmission. These formats, discussed in greater detail below, are important witnesses to other intended audiences and uses of the Etymologiae than the Big Isidores.
How important the small Isidores may have been in the early Middle Ages is attested by the fact that only 108 of the 443 items in the corpus, or approximately a quarter of surviving early medieval manuscripts transmitting the Etymologiae, are manuscripts of the Big Isidore (see Fig. 1). To be more precise, 84 of the Big Isidores are fully preserved or damaged but otherwise well-preserved manuscripts, and 24 survive only as fragments. By contrast, 300 fully preserved or damaged but otherwise well-preserved manuscripts and 21 fragments, or a total of 321 items, can be identified as small Isidores. In the case of nine fragments, it is impossible to determine whether they come from a manuscript of a Big Isidore or a small Isidore, and five items represent manuscripts in which material from the Etymologiae appears as additions in the margins or on blank pages. In addition, three manuscripts contain both the canonical Etymologiae and a selection from this work and were thus counted in both categories.
Fig. 1: Distribution of different manuscript types among the surviving pre-1000 manuscripts of the Etymologiae
It should be stressed that the ratio of roughly 3:1 between small Isidores and Big Isidores revealed in Fig. 1 pertains to the manuscripts transmitting the Etymologiae as they survive from the early Middle Ages, rather than as they were produced in this period. It does not accurately reflect the ratio at which different formats of the Etymologiae circulated, due to the distinct survival rates of manuscripts of different types, as can be demonstrated by splitting the fully preserved and well-preserved manuscripts from fragments (Fig. 2).
Fig. 2: Distribution of different manuscript types among the fully preserved or well-preserved manuscripts and fragments
The splitting reveals extremely uneven rates of survival of Big Isidore among fully preserved/well-preserved manuscripts and fragments. Manuscripts of the canonical Etymologiae amount to only 22% of the former but a staggering 44% of the latter, so that they appear at twice the rate among fragments as among fully preserved manuscripts (and this is a conservative estimate, given that 17% of fragments cannot be assigned to either category). It is clear that the high proportion of Big Isidores among fragments is not a reflection of the high overall proportion of Big Isidores among manuscripts of the Etymologiae produced in the early Middle Ages, but rather a result of the high rate at which Big Isidores were reused in bindings (and the low rate at which the same was true for small Isidores). Such a high rate of survival in maculature is to be expected, especially as the large format and in many cases also high quality of parchment would make Big Isidores particularly suitable for recycling.
Material criteria affected not only the rates of reuse for binding purposes but also played a crucial role in the overall survival rates of early medieval manuscripts. Size, script, the quality of execution, the status and history attributed to particular books, but also whether they were library books or not, whether they were carried around or remained mostly at a single location, whether they were bound or remained in the state of loose quires, and whether they were subjected to intense use all affected the chances of survival of manuscripts. As it happens, criteria positively affecting survival strongly correlate with the material properties of Big Isidores. It is, thus, certain that the ratio of 22% of Big Isidores among fully preserved/well-preserved manuscripts also does not reflect the production rates accurately. In reality, they probably never amounted to more than 15-20% of the early medieval population of manuscripts transmitting the Etymologiae and possibly constituted less than 10% of this population. Thus, while the Big Isidore was an important medieval format of the Etymologiae – the only format that goes back directly to Isidore and his first editor, Braulio of Zaragoza – it was also restricted to particular groups of users and contexts of use. The text of the Etymologiae circulated more widely selectively through the medium of small Isidores, which were the main vehicle of the reception of this work in the early Middle Ages.
Some basic facts about the oldest manuscripts of the Etymologiae
The uneven survival rates of different kinds of manuscripts affect not only the relative ratio of the Big Isidores and small Isidores in the corpus examined here. As is well known, few to no codices survive from certain periods and regions, with most pre-1000 manuscripts being the products of Carolingian scriptoria, whose prodigious output overshadows any other regional production and diffusion trends in our evidence. The surviving early medieval manuscripts of the Etymologiae reflect this Carolingian supremacy, with more than 70% of them copied in a form of Caroline minuscule (see Fig. 3 and Tab. 2). Only five fully preserved manuscripts transmitting the Etymologiae copied in insular scripts survive. However, there are at least 14 insular fragments, including remnants of seven Big Isidores dismembered at continental centers with insular foundations. By contrast, the 16 known manuscripts copied in Visigothic minuscule include no fragments and are divided equally between Big and small Isidores. All other scripts are represented in the examined corpus only marginally.
Fig. 3: Distribution of various scripts among the surviving pre-1000 manuscripts containing the Etymologiae
|script||Big Isidores||small Isidores||unknown|
Tab. 2: Distribution of various scripts among the surviving pre-1000 manuscripts of the Etymologiae
As to the chronological distribution of the surviving pre-1000 manuscripts of the Etymologiae, more than two-thirds of the identified codices can be dated to the ninth century (see Fig. 4 and Tab. 3). This is certainly due to the fact that we lack a resource comparable to Lowe’s CLA or Bischoff’s Katalog for the tenth and eleventh centuries, so that quite a few post-900 small Isidores and fragments escape our attention. However, it is evident that these ratios also reflect the tremendous productivity of Carolingian scriptoria in the ninth century. More than twice as many Big Isidores, which we should assume have been identified satisfactorily, survive from the ninth century as from the tenth and beginning of the eleventh centuries combined.
Fig. 4: Chronological distribution of the surviving pre-1000 manuscripts containing the Etymologiae
|century||Big Isidores||small Isidores||unknown|
Tab. 3: A chronological overview of the surviving pre-1000 manuscripts of the Etymologiae
It is no surprise that almost half of the surviving pre-1000 manuscripts of the Etymologiae were produced in France, mostly in the north (see Fig. 5 and Tab. 4). German centers account for little over half of the surviving production from French scriptoria and Italian scriptoria for about half of the surviving German production (or one-quarter of the French). In the end, almost 90% of surviving and identified pre-1000 manuscripts transmitting the Etymologiae can be assigned to one of these three regions. However, this distribution ratio of 2:1:0.5 may be illusory, as the manuscripts of Big Isidore show. Almost as many Italian manuscripts of the canonical Etymologiae survive as German ones, and both areas more closely match the number of Big Isidores to survive from French scriptoria (this distribution ratio of Big Isidores is, in other words, closer to 2:1:1).
Fig. 5: Distribution of the surviving pre-1000 manuscript of the Etymologiae based on the region of production
|place||Big Isidores||small Isidores||unknown|
Tab. 4: An overview of the surviving pre-1000 manuscripts of the Etymologiae by region of production
The comparison of these two ratios suggests what is missing in the picture: the Italian manuscripts, principally those that transmitted the Etymologiae selectively, as well as fragments. Either the small Isidores from Italy survived at rates significantly lower than those from German and French scriptoria or they remain unidentified. The latter is certainly true to some extent, given the current state of cataloguing at the libraries most likely to own early medieval Italian codices, namely Italian ones, including the Biblioteca Apostolica Vaticana. The uneven numbers of identified early medieval manuscripts transmitting the Etymologiae in the collections of the Biblioteca Apostolica Vaticana (40 items) and the Bibliothèque Nationale de France (88 items) is revelatory. It implies that there are likely many unidentified manuscripts containing the Etymologiae in the former institution. One can guess that perhaps as many as 40 early medieval codices and fragments of the Etymologiae currently held by institutions on the Apennine peninsula still await identification. Italian libraries thus seem to be a fertile ground for hunting early medieval Isidores.
To the extent the surviving early medieval manuscripts of the Etymologiae can be assigned to particular centers of manuscript production, several well-known scriptoria emerge as focal points of the copying and, as a result, also of diffusion and appropriation, of the Etymologiae in Carolingian period. At the top of the list is St. Gallen, which was involved in the production of 24 items from the corpus, including six Big Isidores, and also the compilation of two early medieval redactions of Isidore’s encyclopedia. It deserves to be considered the foremost hub of Carolingian Isidorian studies. St. Gallen is followed by Reims (13 items, including three Big Isidores), Tours (12 items, including two Big Isidores), Corbie (nine items, including three Big Isidores), Freising (nine items, including one Big Isidore), Verona (eight items), and Mainz (eight items, including four Big Isidores). These numbers also reveal that large centers with well-organized scriptoria may have had an interest in owning multiple Big Isidores and also that they may have substantially differed in their strategies of copying. Some centers clearly preferred to produce or amass Big Isidores, while others owned just one copy of the canonical Etymologiae, showing more interest in owning particular types of small Isidores.
The anatomy of small Isidores: what was selected and in what context?
Already Charles H. Beeson, who published the first survey of the oldest manuscripts of Isidore’s works in 1913, noted that material from some of the 20 books of the Etymologiae was transmitted selectively much more frequently than from others. Indeed, as the numbers of manuscripts of the small-Isidore type containing passages from particular books (Fig. 6) indicate, material from the first decade of the Etymologiae appears in small Isidores significantly more often than material from the second decade. To be more precise, except for books IV (on medicine) and X (an alphabetical glossary), each of the books in the first decade of the Etymologiae appears more frequently in the small Isidore format than any of the books of the second decade. Book VI, which deals with the Bible, books and libraries, Church councils, Easter and other holidays, and Church rites, seems to have been excerpted most commonly (76 times, or in more than a quarter of the surviving small Isidores), followed by books I (De grammatica, 64 times, or in more than one-fifth of the surviving small Isidores), V (De legibus et temporibus, 55 times), VII (De Deo, angelis et sanctis, 54 times), and IX (De linguis, gentibus, regnis, militia, civibus et affinitatibus, 53 times). On the opposite end of the scale are books XII (De animalibus, six times), XVII (De rebus rusticis, ten times), XIX (De navibus, aedificiis et vestibus, eight times), and XX (De domo et instrumentis domesticis, six times).
Fig. 6: Number of surviving pre-1000 manuscripts containing excerpts from individual books of the Etymologiae
There exist manifest differences between the manner and context in which material from individual books of the Etymologiae was reused. Material from some books appears, for example, always or very frequently in the company of bits and pieces from other books, while material from other books is commonly transmitted on its own. The difference in this regard is particularly evident between books I and VI, the two books most frequently appearing in non-canonical contexts. Almost two-thirds of the 64 manuscripts containing passages from book I do not contain material from any other book of the Etymologiae. Moreover, more than 40% of them transmit large segments of or the entire book I. By contrast, material from book VI commonly appears in combination with material from other books, especially books VII, VIII, and IX. Unsurprisingly, this thematically diverse book is rarely if ever transmitted as a whole, but rather its chapters or even smaller units were repurposed separately. Material from the medical book IV appears in isolation even more commonly than the material from book I (in more than 80% of cases), and this book, too, is prone to appear in its entirety. By contrast, material from books XI-XX seldom appears on its own, being chiefly transmitted in combination with material from other books.
While small Isidores are structurally complex, they can be classified into seven useful categories: 1) a single excerpt (any material corresponding to a chapter or less transmitted alone); 2) multiple excerpts (disjoined excerpts that appear in different parts of a manuscript); 3) a set of excerpts (excerpts adjoined together into a larger unit smaller than a book section); 4) a book section (one or more thematically coherent units of the Etymologiae delineated by medieval tituli smaller than a book but larger than a set of excerpts, e.g., the first or the second part of book V); 5) a book or a set of books (one of the 20 books of the Etymologiae or several books joined together); 6) an excerpt collection (excerpts joined together into a larger unit, clearly recognized as a separate but coherent entity, e.g., having a title); and 7) an epitome (a selection of chapters that covers the entire range of the Etymologiae, often produced with the explicit intent of abbreviation). The following doughnut chart (Fig. 7) shows the relative distribution of these structural subtypes in the corpus. The inner segment represents the total number of manuscripts of small Isidores (328 mss.), as a result of assigning each manuscript to only one of the seven categories based on the unit of the highest order present (i.e., if a manuscript contains both a book and a single excerpt, it is classified under “book” only). The outer doughnut corresponds to values when all units are counted separately, such that several manuscripts are counted in twice or three times (379 entities).
Fig. 7: The relative distribution of different structural subtypes among the surviving early medieval manuscripts of the small Isidore type
Fig. 7 shows that isolated excerpts represent the most common kind of selective transmission of the Etymologiae (26-27% of small Isidores). Many of the manuscripts transmitting only one excerpt from the Etymologiae can be characterized as miscellanies or handbooks with diverse content (~30%), but there are also quite a few canon law collections (~23%) and grammatical compendia (~10%), due to the habit of attaching certain chapters of Isidore’s knowledge corpus to these collections (e.g., De canonibus conciliorum from book VI and De clericis from book VII to the former, and De orthographia from book I to the latter). Sets of excerpts (21-22% of small Isidores) likewise tend to appear in miscellanies and handbooks of diverse content (~25%) and to a lesser extent in legal manuscripts (~15%) as a result of the fact that they often feature the sequence of chapters on consanguinity from book IX. Multiple excerpts (10-12% of small Isidores) appear likewise in miscellanies and handbooks with diverse content (~39%) and to a lesser extent in computistic collections (~18%), which feature segmented material from books III, V, VI, and XIII. Counted together, these three categories of transmission amount to around 57-61% of all manuscripts classified as small Isidores. They do not reveal any strong correlation between excerption and particular contexts of use, as the only type of books that stand out as associated with them are miscellanies, that is, books lacking a clear thematic focus. While excerpts from the Etymologiae do appear overall more frequently in compendia and collections, that is, books having a thematic focus, they appear across many different types of such books, so that they do not create the impression of particular patterns of appropriation or reuse.
This is in contrast to the 35-40% of small Isidores comprising excerpt collections, and separately transmitted books, book sections, and book sequences. Miscellanies feature only marginally as books transmitting these types of non-canonical Etymologiae. Rather, one can see a close alignment between the types of books, the transmission format, and the thematic focus. Also, one can note that some, but not all, books were transmitted separately or compiled into excerpt collections, an indication that this manner of appropriation of the Etymologiae was driven by specific trends. Thus, both parts of book V (De legibus corresponding to chapters 1-27 and De temporibus corresponding to chapters 28-39) were often transmitted in their own right, the former in legal collections (16 manuscripts, ~39% of book sections), and the latter in computistic and quadrivial compendia (eight manuscripts, ~18% of book sections). However, the two sections of book II (De rhetorica corresponding to chapters 1-21 and De dialectica corresponding to chapters 22-31) were hardly ever repurposed in a comparable fashion. The only other book whose sections were transmitted separately in the early Middle Ages was book III dedicated to the quadrivium. Similarly, while in theory most of the twenty books of the Etymologiae could be transmitted separately, this was the fate of only three: the grammatical book I (16 manuscripts, ~52% of mss. separately transmitting books of the Etymologiae), which can be found in grammatical compendia; the medical book IV (seven manuscripts, ~22% of mss. separately transmitting books of the Etymologiae), which appears in medical collections; and the glossary-like book X (5 manuscripts, ~15% of mss. separately transmitting books of the Etymologiae), which made inroads into collections of glossaries. Excerpt collections draw overwhelmingly on books V to IX and typically appear in pastoral collections (~37% of excerpt collections) and, to a lesser extent, in legal collections (~20% of excerpt collections).
Another dimension along which small Isidores can be analyzed is the co-occurrence of material from the individual books of the Etymologiae in the corpus manuscripts. The following graph (Fig. 8) illustrates the strongest ties of co-occurrence. It reveals that most of the books of the second decade, as well as the medical books IV and the glossographic book X, hardly ever appear in combination with other books. By contrast, material from books V, VI, VII, VIII, and IX frequently appear together. Material from book VI accompanies material from book VII in 26 manuscripts, material from book VII is combined with material from book VIII in 29 manuscripts, material from book VI is placed next to material from book VIII in 21 manuscripts, and all three books dealing with God, the Church, and its rites appear together in 19 manuscripts. Book XV is associated with this cluster because its chapter 6 (De aedificiis sacris) was often seen as usefully complementing the books dealing with matters relevant to pastoral care and the education of clergy. Indeed, Fig. 8 reveals the extent to which theological and catechetical concerns drove the repurposing of material from the Etymologiae in the early Middle Ages. This powerful trend can be underestimated because it is realized in many unique handbooks, each representing a different selection of material from the theological books VI, VII and VIII and from other books of the Etymologiae, arranged in a distinct order, and combining the material from Isidore’s knowledge collection with different authors and works. These handbooks are both very different and very similar. They cannot be compared in their particular content, nor should they be expected to feature the exact same sequence of items, even though that happens in some cases. Rather, they are comparable in their function, users, and even physical properties. In many of these handbooks, material from books VI, VII, and VIII fed excerpt collections. The cluster of strongest ties in Fig. 8, thus, overlaps with this category in Fig. 7.
Fig. 8: The extent of co-occurrence of particular books of the Etymologiae in the surviving pre-1000 manuscripts transmitting the text selectively
Fig. 8 also shows a strong tie between books V and IX, or, to be more precise, between the first part of book V on law (De legibus) and the chapters of book IX dealing with the political organization of the state (9.3: De regnis militiaeque vocabulis, and 9.4: De civibus) on the one hand, and the chapters dealing with family ties and consanguinity (9.5: De affinitatibus et gradibus, and 9.6: De agnatis et cognatis) on the other. The former appears next to one or both of the latter in 19 manuscripts. Selected chapters from these two books features prominently in early medieval manuscripts containing legal texts, especially the Breviarium Alarici. Other relatively strong ties exist between books dealing with time-reckoning and heavenly phenomena, especially between the fourth part of book III (De astronomia) and book XIII dealing with the natural world, and between the second part of book V on time-reckoning (De temporibus) and chapter 17 of book VI containing an Easter table. These co-occurrences certainly reflect the repurposing of the Etymologiae in computistic manuals. Finally, one strong trend concerns a lack of co-occurrence. The grammatical book I rarely appears combined with other books. The 42 small Isidores that transmit parts of this book without containing any other material from the Etymologiae are more numerous than any other instance of co-occurrence or isolated transmission in my corpus.
The separate transmission of book I in a grammatical context; the appropriation of parts of books III, V, VI, and XIII in the context of computus and related scientific pursuits; the combined transmission of sections of books V and IX in legal collections; and the compilation of excerpt collections containing material from books VI, VII, and VIII that appear in the context of pastoral care and clerical training represent the most important and the most homogeneous types of early medieval small Isidores in my corpus. They should be seriously considered as rivaling the Big Isidores in their influence and extent of diffusion, especially in certain milieus and among certain classes of readers. While the numbers of surviving manuscripts representing these four strains of selective transmission of the Etymologiae is lower than the numbers of surviving Big Isidores, they are by no means low for an early medieval text. In the light of uneven survival rates of Big and small Isidores outlined above, furthermore, there is a good reason to think that in particular what may be termed the Isidorus grammaticus and the Isidorus pastoralis existed in substantially larger numbers in the early Middle Ages than they do today and may have circulated in more copies than the canonical Etymologiae. These, rather than the canonical Etymologiae, would have been the format in which young oblates and semi-literate priests-in-training were introduced to the great Spaniard. For many of these scolastici and clerici, this may have also been the only manner in which they accessed the Etymologiae.
How we have discussed the various patterns of repurposing and appropriating the Etymologiae in the early Middle Ages may have created the false impression that we were discussing attitudes that were uniformly widespread in the period. This is certainly not the case. These are almost always responses to specific historical conditions, even though the exact context of their origin may be difficult to pinpoint. We can, for example, be reasonably sure that the computistic strain of small Isidores is to be traced to seventh-century Ireland and the specialist computistic interests of the Irish in this period, even though the computistic collections heavily featuring Isidore retained sufficient relevancy in the following centuries to keep on being copied and read in the Carolingian zone. By contrast, the inclusion of the separated book I into grammatical compendia, signaling its introduction into the classroom, is largely a Carolingian phenomenon. This is not because the manuscript evidence for this mode of transmission is exclusively Carolingian, but because the extant evidence from other regions – insular, Visigothic, and Beneventan – suggests a more limited appeal of book I of the Etymologiae as an ars grammatica. Moreover, in Carolingian environment the use of the Isidorus grammaticus reached an unprecedented level that cannot be seen simply as a continuation of earlier trends. The Carolingian collections of the Isidorus pastoralis may have been dependent on older florilegia and collections to a more significant extent, but once again, their extensive popularity in the Carolingian environment and the fact that new collections were produced in the ninth century also identify this strain of small Isidores as having a particular Carolingian quality.
The two most influential strains of Carolingian small Isidores, which both appear in the manuscript evidence at the end of the eighth century and spread rapidly in the following decades, mirror concerns and interests that were also expressed by the Carolingian reform movement. The Isidorus grammaticus and the Isidorus pastoralis, in fact, look markedly like what one could expect to happen to Isidore’s encyclopedia once set on a collision course with the reform movement. Above all, these two types of small Isidores reveal an important aspect of the Carolingian approach to the Etymologiae, namely that for Carolingian users, as is demonstrated by the Isidorus grammaticus and the Isidorus pastoralis but also other forms of Carolingian appropriation of Isidore’s magnum opus, it was primarily an educational resource and not a scholarly encyclopedia akin to Martianus Capella’s De nuptiis Mercurii et philologiae or Pliny’s Historia naturalis. This conveniently pre-packaged knowledge corpus was tapped by the two key projects of the reformers, the establishment of schools where scolastici could be trained in using Latin correctly (hence the appearance of book I of the Etymologiae in Carolingian model grammar books) and the broader education of clergy, which would in turn cultivate the populus Christianus (hence the influx of Isidorian collections into handbooks for bishops and priests). It is not accidental that books of both types containing material from the Etymologiae seem to have been produced mostly in the region of modern France, the heart of the reform movement. For both purposes, the reworking of the material from the Etymologiae involved at least some degree of simplification, including via selection and reassembling, as well as by the addition of explanatory glosses (book I) and the recasting into a dialogue form (books V-IX). For this reason, too, it can be argued that the first half of the Etymologiae, which covers subjects aligned with early medieval education (Liberal Arts, medicine, law, time-reckoning, theology), attained significantly more attention in the Frankish environment than the second half which discusses topics that were more properly encyclopedic and scholarly insofar as they could not be easily integrated into a standard school curriculum.
Nevertheless, the reception of the Etymologiae in the Carolingian period had also its scholarly side that is embodied in the Big Isidores. The character of the appropriation of this type of the Etymologiae, such as the production of new Carolingian redactions of the encyclopedia, the addition of complementary material from authoritative sources, and filling the gaps, can be seen as an attempt to upgrade this certainly advanced but not entirely specialist text emanating from scholarly milieu, and contrasting with the simplifying tendency visible in many small Isidores. Unless perhaps the very act of copying of a Big Isidore should be interpreted as an ambition towards scholarly study, the traces of scholarly engagement with Big Isidores are distributed unevenly across the Carolingian world. The most important scholarly projects concerning the Etymologiae, as far as they can be identified today, seem to have taken place outside the Frankish heartland. Northern Italy, in particular, emerges as a region in which the Etymologiae may have been treated as a scholarly encyclopedia more often than elsewhere. At St. Gallen, presumably influenced by trends from across the Alps, the Etymologiae was also treated more like an encyclopedia than as an educational resource. Manuscripts from both northern Italy and in St. Gallen show traces of engagement with the second half of the Etymologiae, the part that provided much interesting information for scholarly minded readers but had limited use for schools. Two early manuscripts of the Etymologiae from St. Gallen, for example, preserve an identical set of annotations to the second half of the Etymologiae which reveal an interest in plant and animal life and supplementing Isidore’s geographical knowledge of the Apennine peninsula, a sure sign of their origin in eighth-century Italy. The most famous set of annotations to the Etymologiae, the scholia Vallicelliana, also originated in Italy, whether they should be attributed to bishop Grauso of Ceneda or Paul the Deacon. Other rich layers of annotations can be found in manuscripts from Ivrea and Cesena. At least one redaction of the Etymologiae, other than the two stemming from St. Gallen, was produced in northern Italy in the ninth century.
If any generalization can be adduced from the oldest surviving manuscripts, it is the impression of an east-west divide in how Isidore’s work was appropriated in Carolingian times. These regional differences may be rooted in how Isidore was read – whether he was seen as a learned author of an encyclopedia, or rather as a collector of useful knowledge intended for educational purposes. It seems that in the eastern parts of the Empire, and most notably in northern Italy, he was more often the former, while in the Empire’s western parts, particularly in northern France, he was more frequently the latter. This east-west divide is, naturally, no set rule but rather a suggestion that needs to be probed further and should be understood as describing a general tendency with many particular localized variations. Manuscripts from the insular milieu, for example, contain annotations of the scholarly type and also reflect in other ways engagement with the Etymologiae that deviates from what was common in the western parts of the Carolingian empire. The sense of regional difference is a matter of a degree and as such can be refined or even challenged if new manuscripts are brought into the picture – for example, if new codices transmitting the Etymologiae selectively are identified or if the holdings of Italian libraries are better examined, or if the corpus assembled here is re-evaluated just as the older corpus assembled by August E. Anspach has been. Most importantly, it is a divide that can be spotted only from above, when a corpus of manuscripts is scrutinized with the eye of an eagle. This view needs the complement of detailed qualitative studies, especially of those manuscripts that are revealed to be notable outliers or the model examples of regional and epochal trends.
This article came into being in the context of the postdoctoral project Innovating Knowledge: Isidore’s “Etymologiae” in the Carolingian Period funded by the Dutch Research Organization (NWO) and carried out between February 2018 and April 2021 at the Huygens ING, an institute of the Dutch Royal Academy of Arts and Sciences in Amsterdam. I would like to thank Jacques Elfassi for his useful comments on an earlier version of this article and Meg Worley and Christie Pavey for checking my English.
 The most important studies on the Etymologiae, its medieval reception and manuscript tradition include August Eduard Anspach, “Das Fortleben Isidors im VII. bis IX. Jahrhundert,” in Miscellanea Isidoriana (Rome: Universita Gregoriana, 1936), 322–56; Walter Porzig, “Die Rezensionen der Etymologiae des Isidorus von Sevilla. Vorbemerkung,” Hermes 72, no. 2 (1937): 129–70; Bernhard Bischoff, “Die europäische Verbreitung der Werke Isidors von Sevilla,” in Isidoriana: collección de estudios sobre Isidoro de Sevilla, ed. Manuel C. Díaz y Díaz (León: Centro de estudios San Isidoro, 1961), 317–44; Jacques Fontaine, “La diffusion de l’oeuvre d’Isidore de Séville dans les scriptoria helvétiques du haut Moyen Âge,” Schweizerische Zeitschrift für Geschichte 12 (1962): 305–22; Marc Reydellet, “La diffusion des Origines d’Isidore de Séville au Haut Moyen âge,” Mélanges d’archéologie et d’histoire 78, no. 2 (1966): 383–437; Manuel C. Díaz y Díaz, Los capítulos sobre los metales en las “Etimologías” de Isidoro de Sevilla (León: Cátedra de San Isidoro, 1970); Manuel C. Díaz y Díaz, “Problemas de algunos manuscritos hispánicos de las Etimologías de Isidoro de Sevilla,” in Festschrift Bernhard Bischoff zu seinem 65. Geburtstag, ed. Johanne Autenrieth and Franz Brunhölzl (Stuttgart: Hiersemann, 1971), 70–80; Jacques Fontaine, Isidore de Séville et la culture classique dans l’Espagne wisigothique, 2nd edn. (Paris: Études augustiniennes, 1983); Ulrich Schindel, “Zur frühen Überlieferungsgeschichte der Etymologiae Isidors von Sevilla,” Studi medievali 29, no. 2 (1988): 587–605; Carmen Codoñer Merino, José Carlos Martín, and María Adelaida Andrés Sanz, “Isidorus Hispalensis Ep.,” in La Trasmissione Dei Testi Latini Del Medioevo/Medieval Texts and Their Transmission, ed. Paulo Chiesa and Lucia Castaldi, vol. 2 (Florence: SISMEL – Edizioni del Galluzzo, 2005), 274–417; Baudouin Van den Abeele, “La tradition manuscrite des Étymologies d’Isidore de Séville,” Cahiers de recherches médiévales et humanistes 16 (2008): 195–205; Carmen Codoñer Merino, “Transmisión y recepción de las ‘Etimologías’,” in Estudios de latín medieval hispánico. Actas del V Congreso Internacional de Latín Medieval Hispánico, ed. José Martínez Gázquez, Óscar de la Cruz Palma, and Cándida Ferrero Hernández (Florence: SISMEL – Edizioni del Galluzzo, 2011), 5–26; Carmen Cardelle de Hartmann, “Uso y recepción de las Etymologiae de Isidoro,” in Wisigothica. After M. C. Díaz y Díaz, ed. Carmen Codoñer Merino and Paulo Farmhouse Alberto, mediEVI 3 (Florence: SISMEL, 2014), 477–502; and Jacques Elfassi, “Isidore of Seville and the Etymologies,” in A Companion to Isidore of Seville, ed. Andy Fear and Jamie Wood (Boston and Leiden: Brill, 2020), 245–78.
 For a brief quantitative overview of the surviving manuscripts by century, see Van den Abeele, “La tradition manuscrite des Étymologies d’Isidore de Séville,” 197–201.
 The light-weight critical edition published by W. M. Lindsay in 1911 remains the standard critical text of the Etymologiae until this day; Wallace Martin Lindsay, Etymologiarum sive Originum libri XX, 2 vols. (Oxford: Clarendon Press, 1911). A work on a new critical edition cum translation of the Etymologiae, each of the twenty books of Isidore’s work edited and published as a separate volume in cooperation with Belles Lettres, has been ongoing under the aegis of CNRS since 1962. Almost sixty years since the inception of this project, certain books have still not appeared, which just illustrates the difficulty of the task. For the plan of this editorial project, see Jacques Fontaine et al., “Compte rendu du colloque isidorien tenu à l’Institut d’études latines de l’Université de Paris le 23 juin 1970,” Revue d’Histoire des Textes 2, no. 1972 (1973): 282–88.
 See for example Jacques Elfassi, “Connaître la bibliothèque pour connaître les sources: Isidore de Séville,” Antiquité tardive 23 (2015): 59–66; and Jacques Elfassi, “Nuevas fuentes en la biblioteca de Isidoro de Sevilla,” in Latinidad Medieval Hispánica: Congreso Internacional de Latín Medieval Hispánico, 2013, La Nucia, Spain, ed. Juan Francisco Mesa Sanz, MediEVI 14 (Florence: SISMEL – Edizioni del Galluzzo, 2017), 107–16.
 Anspach’s life and scholarly career are detailed in Ángel Custodio Vega, “El Prof. Doctor August Eduard Anspach: semblanza literaria,” in Catálogo de los materiales codicológicos y bibliográficos del legado científico del Prof. Dr. August Eduard Anspach, ed. José María Fernández Catón (León: Centro de estudios e investigación “San Isidoro,” 1966), 9–28; and Theodor Kurrus, “In memoriam: August Eduard Anspach (1860-1943),” in Spanische Forschungen der Görres-Gesellschaft 1, ed. Johannes Vincke, Gesammelte Aufsätze zur Kulturgeschichte Spaniens 26 (Münster: Aschendorff, 1971), 349–56.
 See footnote 3.
 August Eduard Anspach, Taionis et Isidori Nova Fragmenta et Opera, Textos Latinos de La Edad Media Española 2 (Madrid: Bermejo, 1930); and Anspach, “Das Fortleben Isidors im VII. bis IX. Jahrhundert.”
 José María Fernández Catón, Catálogo de los materiales codicológicos y bibliográficos del legado científico del Prof. Dr. August Eduard Anspach (León: Centro de estudios e investigación “San Isidoro,” 1966).
 José María Fernández Catón, Las “Etimologías” en la tradición manuscrita medieval estudiada por el Prof. Dr. Anspach (León: Centro de estudios e investigación “San Isidoro,” 1966).
 This investigation has been carried out in the context of the Innovating Knowledge: Isidore’s “Etymologiae” in Carolingian period project. A 1000-word summary of this project can be found at: https://mittelalter.hypotheses.org/21234.
 Van den Abeele, “La tradition manuscrite des Étymologies d’Isidore de Séville,” 198; and Cardelle de Hartmann, “Uso y recepción de las Etymologiae de Isidoro,” 477.
 This chart produced on the basis of Anspach’s handlist differs slightly from the chart presented in Van den Abeele, “La tradition manuscrite des Étymologies d’Isidore de Séville,” 199. The differences are due to our distinct interpretation of Anspach’s manuscript items.
 The problems with the handlist are also discussed in Van den Abeele, 198.
 Items no. 23 and 40, for example, both describe the excerpt of Etym. 1.6-14 in St. Gallen, Stiftsbibliothek, MS 876.
 Thus, Anspach included the fourteenth-century Paris, BnF, Lat. 7418, among ninth-century codices of the Etymologiae (no. 148 in his handlist), but assigned the ninth-century Paris, BnF, Lat. 7587 to the eleventh century (no. 318). Predating of post-1000 manuscripts concerns 15 items in Anspach’s/Catón’s handlist. Postdating affected 8 items.
 For example, Isidore is mentioned in the title of an excerpt on fols. 123v-124v in Avranches, Bibliotheque municipale, MS 109 (no. 224), but this text is not from the Etymologiae. Similarly, Anspach indicates that Rome, Biblioteca Casanatense, MS 54 (no. 278) contains excerpts from book III of the Etymologiae, but these excerpts bearing Isidore’s name come rather from Bede’s De natura rerum.
 For example, Bern, Burgerbibliothek, MS 123 (no. 229) is a manuscript containing Ars Bernensis, a grammatical work citing Isidore, rather than excerpts from book I of the Etymologiae. The fragment Paris, BnF, Lat. 7491 (fols. 118-137) is not a fragment of book I of the Etymologiae but of Cruindmel’s Ars metrica, a text dependent on the Etymologiae. Misidentification concerns altogether 42 items in Anspach’s handlist.
 This is Anspach’s items no. 263 (Metz, which was assigned to the tenth century and should contain the first three books of the Etymologiae together with Alexandri opus de historia naturali astronomiae).
 These are items no. 78 (Wiesbaden, Landesbibliothek, MS 242), 104 (Chartres, Bibliothèque municipale, MS 80), 105 (Chartres, Bibliothèque municipale, MS 92), 187 (Chartres, Bibliothèque municipale, MS 68), 201 (Metz, Bibliothèque municipale, MS 179), 239 (Chartres, Bibliothèque municipale, MS 63), 264 (Metz, Bibliothèque municipale, MS 145), and 355 (Strasbourg, Bibliothèque municipale, C IV 15).
 Elias Avery Lowe, Codices Latini Antiquiores: A Palaeographical Guide to Latin Manuscripts Prior to the Ninth Century, 11 vols. (Oxford: Clarendon Press, 1934-66).
 Bernhard Bischoff, Katalog der festländischen Handschriften des neunten Jahrhunderts: (mit Ausnahme der wisigotischen), ed. Birgit Ebersperger, 4 vols., Veröffentlichungen der Kommission für die Herausgabe der mittelalterlichen Bibliothekskataloge Deutschlands und der Schweiz (Wiesbaden: Harrassowitz, 1998-2017). The CLA and Bischoff’s Katalog serve as a source of most of the dates and places of production invoked in this article.
 Karl Forstner was, for example, able to discover several fragments of a large and important codex of the Etymologiae copied at the turn of the ninth century in Salzburg that were unknown to Bischoff or the Codices latini antiquiores; see Karl Forstner, “Ergänzungen zu B. Bischoffs Hss Katalog (Salzburger Fragmente),” Scriptorium 62 (2008): 122–38. Images of a ninth-century fragment of the Etymologiae unknown to Bischoff has been recently published in Dalibor Havel, Počátky latinské písemné kultury v českých zemích: nejstarší latinské rukopisy a zlomky v Čechách a na Morave, Opera Facultatis Philosophicae Universitatis Masarykianae 479 (Brno: MUNI, 2018), 141–44. Another early medieval fragment of the Etymologiae with Bohemian provenance unknown to Bischoff is today located in the diocesan archive in León as a part of Anspach’s estate. Finally, a substantial set of ninth-century fragments of the Etymologiae have been recently described in Esther van de Vrie, “Een tekst voor studie, toen en nu,” in Perkament in stukken: Teruggewonden middeleeuwse handschriftfragmenten, ed. Bart Jaski, Marco Mostert, and Kaj van Vliet (Hilversum: Verloren, 2018), 168–71.
 Although, occasionally, a manuscript of the canonical Etymologiae also eludes scholars, as has been the case with a tenth-century codex of Isidore’s work owned by Sir Thomas Phillipps (his no. 2129) that went missing already in the nineteenth century and resurfaced at an auction in Christie’s in 2010. See the description at: https://www.christies.com/lotfinder/books-manuscripts/isidore-of-seville-etymologiae-books-i-xi-5370909-details.aspx?from=salesummery&intobjectid=5370909&sid=5fa6d866-100c-41b7-abd0-01de6f8a084c. I want to thank Jacques Elfassi for sharing with me a draft of his article on this codex, which he was able to examine in 2014.
 A complete overview of these 443 manuscripts with their descriptions can be found in the online database of the Innovating Knowledge project, at: etymologiae.ms. This number increases Anspach’s count by almost one third. If it is used as a correction coefficient of his total number of medieval manuscripts of the Etymologiae, we should think that there are almost 1,800 surviving codices transmitting this text.
 See Bischoff, “Die europäische Verbreitung der Werke Isidors von Sevilla,” 339. To provide several examples, Theodulf of Orleans used sections of book VI as prefaces in his revision of the Bible, Etym. 6.16 (De canonibus conciliorum) was used as a preface to the Dionysio-Hadriana and several other collections of canon law, Etym. 1.22 (De notis vulgaribus) appears as a preface in at least two Carolingian copies of the Commentarii notarum Tironianum, the first section of book III on quadrivial disciplines dedicated to arithmetic was attached to Boethius’s Institutio arithmetica, and the third section of the same book concerned with music appears in the same position in one Carolingian manuscript of Boethius’s Institutio musica and in one Carolingian manuscripts of Augustine’s Musica.
 At least six ninth- and tenth-century manuscripts, for example, transmit a collection called De legibus divinis sive humanis that combines excerpts from books II, V and XVIII of the Etymologiae; see Ernest-Joseph Tardif, “Un abrégé juridique des Étymologies d’Isidore de Séville,” in Mélanges Julien Havet (Paris: Ernest Leroux, 1895), 659–81. At least nine manuscripts from the ninth and tenth century contain a different nameless question-and-answer collection described in Michael Gorman, “The Carolingian Miscellany of Exegetical Texts in Albi 39 and Paris Lat. 2175,” Scriptorium 51 (1997): 336–54. Yet another important Isidorian excerpt collection was incorporated into the Collectio Sangermanensis; see Michael Stadelmaier, Die Collectio Sangermanensis XXI titulorum: eine systematische Kanonessammlung der frühen Karolingerzeit, Freiburger Beiträge zur mittelalterlichen Geschichte 16 (Frankfurt am Main: Peter Lang, 2004). A fourth collection known from at least two ninth-century manuscripts and additional post-1000 codices has been analyzed in María Adelaida Andrés Sanz, “Una reelaboración de textos isidorianos en forma de interrogationes et responsiones,” Helmantica 57, no. 172 (2006): 29–48. Other collections are discussed in Cardelle de Hartmann, “Uso y recepción de las Etymologiae de Isidoro,” 499.
 In Trier, Stadtbibliothek, MS 40/1018 8O, a collection of glossaries, excerpts from book IV of the Etymologiae on medicine appear copied in the margins of the first folia. In Cologne, Dombibliothek, MS 123, a collection of canon law, a small snippet of book I appears added in a blank space on fol. 79v. The margins of a Psalter, Oxford, Bodleian Library, Laud Lat. 35, contain the opening of book I of the Etymologiae. The famous notebook of Grimald of St. Gallen, St. Gallen, Stiftsbibliothek, MS 397 contains many contemporary additions on blank leaves, among which are excerpts from books II, III, V, and VI of the Etymologiae.
 Codoñer Merino, “Transmisión y recepción de las ‘Etimologías’,” 5–7.
 That the ninth century may have been the peak of the phenomenon of repurposing and appropriation was suggested in Van den Abeele, “La tradition manuscrite des Étymologies d’Isidore de Séville,” 200–01. However, Van den Abeele bases himself on numbers of manuscripts derived from Anspach, which cannot be used to draw an accurate comparison between different periods of the Middle Ages.
 A useful overview of these strategies is outlined in Cardelle de Hartmann, “Uso y recepción de las Etymologiae de Isidoro.”
 However, even editors have good reasons not to neglect small Isidores, for it is clear that the two formats interbred and that material passed from one format to another and back. The first book of a copy of the entire Etymologiae, Zofingen, Stadtbibliothek, Pa 32 (9th c., 2/2, St. Gallen), was, for example, almost certainly copied from a handbook containing book I of the Etymologiae as a self-standing text; see Evina Steinová, “Two Carolingian Redactions of the Etymologiae from St. Gallen,” Mittellateinisches Jahrbuch, forthcoming. Similarly, the additions from the De natura rerum in book III of the Etymologiae may be a debt of Irish computistic compendia. The transfer of material from Big Isidores to small Isidores was even more common, which makes them valuable as reflecting text of manuscripts that were lost.
 The oldest stages of the diffusion of the Etymologiae and the problems surrounding the reconstruction of Isidore’s version of his text and of Braulio’s involvement have been discussed in innumerable scholarly articles, yet many of the most pressing questions have not been settled satisfactorily. It is unclear how Braulio’s Renotatio and the remarks about the text of the Etymologiae in his letters to Isidore should be understood, whether Isidore produced several versions of the text that circulated before his death, what was the potential relationship of these versions, and how they relate to the surviving early medieval manuscripts. A summary treatment of the issues is provided in Codoñer Merino, Martín, and Andrés Sanz, “Isidorus Hispalensis Ep.,” 281–86.
 See Evina Steinová, “The Materiality of Innovation: Formats and Dimensions of the Etymologiae of Isidore of Seville in the Early Middle Ages,” in Entangled Manuscripts: 600-1200, ed. Anna Dorofeeva and Michael J. Kelly (Binghamton: Gracchi Books, forthcoming).
 Although already one of the oldest surviving manuscripts of the canonical Etymologiae, Milan, Biblioteca Ambrosiana, L 99 sup. (8th c., 2/2, Bobbio) is a remnant of a two-volume copy, the division into two halves is probably an innovation of a text originally designed to be copied in a single volume. Apart from the description of the codex nimiae magnitudinis in Braulio’s Renotatio, the chief reasons to think so is that the two-volume and multi-volume copies remained in minority throughout the early Middle Ages.
 The average page height of these manuscripts is approximately 310 mm (12.20 in.) and the average page width is approximately 225 mm (8.86 in.), meaning that their average taille (height + width) is 535 mm (21.06 in.). This is 67 mm (2.64 in.) more than the average taille of an eighth-century Latin manuscript (469 mm [18.46 in.]) and 47 mm (1.85 in.) more than the average taille of a ninth-century Latin manuscript (488 mm [19.21 in.]); see Marilena Maniaci, “Costruzione e gestione dello spazio scritto fra Oriente e Occidente: principi generali e soluzioni specifiche,” in Scrivere e leggere nell’alto Medioevo, Settimane di studio del Centro italiano di studi sull’alto Medioevo 59 (Spoleto: Centro Italiano di Studi sull’Alto Medioevo, 2012), 484.
 See George D. Greenia, “The Bigger the Book: On Oversize Medieval Manuscripts,” Revue Belge de Philologie et d’histoire 83 (2005): 723–46.
 The Tours Big Isidore (Tours, Bibliothèque municipale, MS 844) is still in Tours, the Einsiedeln Big Isidore (Einsiedeln, Stiftsbibliothek, MS 167) in Einsiedeln, the Laonese Big Isidore (Laon, Bibliothèque municipale, MS 447) is still in Laon, the Freising Big Isidore (Munich, Bayerische Staatsbibliothek, Clm 6250) was moved to Munich only in the nineteenth century, two of three Reims Big Isidores (Reims, Bibliothèque municipale, MS 425 and 426) are still in Reims and five of six Big Isidores produced at St. Gallen (St. Gallen, Stiftsbibliothek, MS 231-232, 233, 235, 236 and 237) in St. Gallen, while the Salzburg Big Isidore, even if dismembered and reused for binding, is still in Salzburg.
 The corpus examined here includes 23 items that fit into the category of “small” manuscripts as defined in Carla Bozzolo and Ezio Ornato, Pour une histoire du livre manuscrit au Moyen Âge: Trois essays de codicologie quantitative, Équipe de recherche sur l’humanisme français. Textes et études 2 (Paris: CNRS, 1980), 218.
 Cardelle de Hartmann, “Uso y recepción de las Etymologiae de Isidoro,” 482.
 These are Berlin, Staatsbibliothek, Hamilton 689 (11th c., perhaps northern Italy), in which the Etymologiae are preceded by other texts including Isidore’s De natura rerum, poems about astronomical phenomena, and an excerpt from the first chapter of book XI of the Etymologiae, Berlin, Staatsbibliothek, Lat. fol. 641 (fols. 17-257) (9th c., med., northern Italy), in which the Etymologiae are followed by a number of grammatical texts including the Ars minor of Donatus with a commentary, glossaries, and excerpts from books I and II of the Etymologiae, and St. Gallen, Stiftsbibliothek, MS 235 (c. 800, St. Gallen), in which the chapter about the etymologia (Etym. 1.29) is used as a general preface to books XII-XX of the Etymologiae.
 See Elisabeth Pellegrin, “Fragments et membra disiecta,” Codicologica 3 (1980): 73.
 Some of these factors are discussed in Bozzolo and Ornato, Pour une histoire du livre manuscrit au Moyen Âge, 72–83.
 See David Ganz, “Book Production in the Carolingian Empire and the Spread of Caroline Minuscule,” in The New Cambridge Medieval History 2: C. 700- c. 900, ed. Rosamond McKitterick (Cambridge: Cambridge University Press, 1995), 786–87.
 More precisely, 295 manuscripts were copied in pure Caroline minuscule, 20 manuscripts in a later form of the script in the state of transition towards the Gothic script, and 9 in an early form of Caroline minuscule.
 These include two Big Isidores, Cambridge, Trinity College, B.15.33 (10th c., Winchester) and Oxford, Queen’s College, MS 320 (10th c., ¾, England), Karlsruhe, Badische Landesbibliothek, Aug. Perg. 167 (c. 848, northeastern France), Kassel, Universitätsbibliothek, 2o Ms. philol. 2 (8th/9th c., perhaps St. Amand), and St. Gallen, Stiftsbibliothek, MS 913 (8th c., 2/2, Germany).
 Three leaves of the old Fulda Big Isidore in Anglo-Saxon minuscule survive as Augsburg, Staats- und Stadtbibliothek, Fragm. lat. 42 and Paris, BnF, Lat. 10403 (fol. 1). The four folia preserved today partially in the Landeskirchliches Archiv in Kassel and partially in Hersfeld are presumably remnants of the old Hersfeld copy. The three folia preserved in the University library in Düsseldorf as well as in the abbey library of Gerleve were produced in Northumbria, but were on the Continent from early on, presumably at Werden. I am currently preparing a separate article about this and other early medieval fragments of the Etymologiae.
 The most remarkable fact about these manuscripts is their size. The four largest manuscripts in my corpus are all Visigothic. The eight Visigothic Big Isidores have the average height of approximately 370 mm [14.57 in.] (about 50 mm [1.97 in.] more than Big Isidores as a group) and the average width of approximately 260 mm [10.24 in.] (about 35 mm [1.38 in.] more than Big Isidores as a group). Even Visigothic manuscripts that do not transmit the canonical Etymologiae are uncharacteristically large: two miscellanies with similar content copied in the second half of the tenth century in northern Spain each have pages measuring more than 450 mm (17.72 in.) in height and at least 300 mm (11.81 in.) in width.
 The absence of tenth-century fragments from the record is particularly noticeable. Of the 54 fragments in the corpus, 7 predate the ninth century, 39 were dated to the ninth century, and only 8 can be dated to the tenth century.
 The third option is that Italian users were less interested in small Isidores than their peers from the German and Frankish areas. It is not entirely unthinkable, especially in light of the evidence mentioned in the conclusion to this article, which shows that in Italy the Etymologiae were read as a scholarly encyclopedia more commonly than anywhere else.
 St. Gallen’s role in copying and disseminating of the Etymologiae is discussed in Steinová, “Two Carolingian Redactions of the Etymologiae from St. Gallen.”
 In this regard, Fleury represents an interesting case of a center to which only one Big Isidore manuscript can be attributed (now surviving as a fragment, Wolfenbüttel, Herzog August Bibliothek, Helmst. 455 [fol. 1]), but which either produced or possessed five collections transmitting material from the first book of the Etymologiae in grammatical context.
 Charles Henry Beeson, Isidor-Studien, Quellen und Untersuchungen zur lateinischen Philologie des Mittelalters, 4.2 (Munich: Beck, 1913), 83.
 Overall, medieval users seem to have treated the first ten and the other ten books differently; see for example Codoñer Merino, “Transmisión y recepción de las ‘Etimologías’,” 11.
 This overview of the selective transmission of material from various books of the Etymologiae is based on a subset of 299 manuscripts from the corpus excluding the canonical Etymologiae, fragments, and marginalia.
 Many scholars have noted that the first book of the Etymologiae was often transmitted as a self-standing grammatical text in grammatical handbooks; see Max Manitius, Geschichte der lateinischen Literatur des Mittelalters, vol. 1 (Munich: Beck, 1911), 67; Louis Holtz, Donat et la tradition de l’enseignement grammatical: étude sur l’Ars Donati et sa diffusion (IV. – IX. siècle) et éd. crit. (Paris: CNRS, 1981), 260; Vivien Law, The Insular Latin Grammarians, Studies in Celtic History 3 (Woodbridge: Boydell Press, 1987), 24; and Codoñer Merino, “Transmisión y recepción de las ‘Etimologías’,” 8. I am currently preparing an article about this Isidorian De grammatica and its manuscript transmission.
 This is particularly true for chapter 16 on Church councils, which appears in manuscripts of canon law from the turn of the ninth century onwards, chapter 17 on the calculation of Easter and its celebration, which appear commonly in computistic manuals, and chapter 19 on Church rites and sacraments that appears in pastoral collections. The oldest surviving manuscripts of canon law containing Etym. 6.16 as a preface are those of the eighth-century Collectio Herovalliana, Paris, BnF, Lat. 2123 (8th/9th c., Flavigny) and Paris, BnF, Lat. 3848B (8th/9th c., Burgundy, prov.: Flavigny); see also Friedrich Maassen, Geschichte der Quellen und der Literatur des Canonischen Rechts im Abendlande bis zum Ausgange des Mittelalters, vol. 1: Die Rechtssammlungen bis zur Mitte des 9. Jahrhunderts (Graz: Leuschner & Lubensky, 1870), 403–04.
 This Isidorus medicus has been examined in Heinz-Albert Schütz, “Die Schrift ‘De medicina’ des Isidor von Sevilla: ein Beitrag zur Medizin im spätantiken Spanien” (Ph.D. diss., University of Giessen, 1984); and Arsenio Ferraces Rodríguez, ed., “Isidorus medicus”. Isidoro de Sevilla y los textos de medicina, Monografias 113 (Coruña: Universidade da Coruña, 2005). The oldest surviving manuscript of this type is Glasgow, University Library, Hunter 96 (8th/9th c., Septimania).
 Material from book XIII, for example, appears in combination with material from books III, V and VI in computistic context, see Immo Warntjes, “Isidore of Seville and the Formation of Medieval Computus,” in A Companion to Isidore of Seville, ed. Andy Fear and Jamie Wood (Boston and Leiden: Brill, 2020), 457-523.
 It may be useful here to clarify that 282 manuscripts fall into one category, 41 manuscripts fall into two categories, and 5 manuscripts fall into three categories.
 The six manuscripts containing epitomes amounting to 2% of small Isidores are not counted in because as abbreviations of the entire Etymlogiae, they do not have a specific thematic focus. To my knowledge only three of them were studied. Vatican, BAV, Vat. Lat. 5764 (9th c., ¼, Verona) is treated in Bischoff, “Die europäische Verbreitung der Werke Isidors von Sevilla,” 339; and especially Codoñer Merino, “Transmisión y recepción de las ‘Etimologías’,” 15–18. The Anglo-Saxon epitome in Paris, BnF, Lat. 1750 (fols. 140-152) (9th c., ½, northern France) has been analyzed in Michael Lapidge, “An Isidorian Epitome from Early Anglo-Saxon England,” Romanobarbarica 10 (89 1988): 443–83. The epitome of the Etymologiae in Leiden, Universiteitsbibliotheek, Voss. Lat. O 15, the famous notebooks of Ademar of Chabannes (11th c., ½, Chabannes and Limoges), has been studied in Ad van Els, “Een leeuw van een handschrift: Ademar van Chabannes en MS Leiden, Universiteitsbibliotheek, Vossianus Latinus Octavo 15” (Ph.D. diss., Utrecht University, 2015).
 The transmission of De legibus is treated in Tardif, “Un abrégé juridique des Étymologies d’Isidore de Séville”; and Rosamond McKitterick, The Carolingians and the Written Word (Cambridge: Cambridge University Press, 1989), 45–55.
 The only two cases known to me are Paris, BnF, Lat. 14116 (10th/11th c., France, prov.: Saint-Maur-des-Fosses), in which De rhetorica appears as an introduction to Cicero’s De rhetorica and Boethius’s De topicis differentiis, and Valenciennes, Bibliothèque municipale, MS 404 (9th c., ¾, France, prov.: St. Amand), in which the entire book II appears as an introduction to Alcuin’s De dialectica.
 Its first two parts (De mathematica, chapters 1-14) feature in quadrivial compendia (6 manuscripts), the third part (De musica, chapters 15-23) can be found as a preface in manuscripts of Boethius’s Institutio musica and in other musical compendia (5 manuscripts), while the fourth part (De astronomia, chapters 24-71) appears in manuscripts with computistic or more broadly scientific content (8 manuscripts).
 The latter book has been studied in Carmen Codoñer Merino, Introducción al Libro X de las “Etymologiae”: su lugar dentro de esta obra, su valor como diccionario (Logroño: Logroño Fundación San Millán de la Cogolla, 2002).
 Compare with Susan Keefe, Water and the Word: Baptism and the Education of the Clergy in the Carolingian Empire, vol. 1, Publications in Mediaeval Studies (Notre Dame: University of Notre Dame Press, 2002), 13.
 The graph includes only cases when material from particular books appear next to each other in at least five manuscripts.
 Compare with Keefe, Water and the Word, 1:137–39.
 See Tardif, “Un abrégé juridique des Étymologies d’Isidore de Séville,” 663. The oldest surviving manuscript transmitting De legibus together with Etym. 9.3-4 is Paris, BnF, Lat. 4403A (8th c., med., northern France), a codex of the Breviarium Alarici.
 In the computistic manuscript Strasbourg, Bibliothèque universitaire, MS 326 (10th c., Limoges or Angoulême), De temporibus is combined with De astronomia, the first eleven chapters of book XIII, and chapters 5 and 6 of book IX on consanguinity into an Isidorian collection on fols. 191r-201v. A similar Isidorian computistic collection consisting from Etym. 6.17, De temporibus, and De astronomia appears on fols. 107v-124v of Milan, Biblioteca Ambrosiana, H 150 inf. (c. 810, France). Yet another collection combining together Etym. 6.17 with De astronomia and the first eleven chapters of book XIII can be found in Berlin, Staatsbibliothek, Phillipps 1831 (9th c., in., Verona). A fuller version of identical collection also containing De temporibus appears in Paris, BnF, Lat. 5239 (10th c., 1/3, Limoges).
 Warntjes, “Isidore of Seville and the Formation of Medieval Computus.”
 See Idem.
 In the insular milieu, book I was used as a basis for many grammatical works; see Law, The Insular Latin Grammarians. On these grounds perhaps, Parkes believed the separate transmission of the first book of the Etymologiae had Irish roots; Malcolm B. Parkes, Pause and Effect: An Introduction to the History of Punctuation in the West (Aldershot: Scolar, 1992), 22–23. However, it can be pointed out that book I was similarly used by Hilderic of Monte Cassino and parts of it appear in the important Monte Cassino compendium of Liberal Arts, Paris, BnF, Lat. 7530; see Anselmo Lentini, Ilderico e la sua Ars grammatica, Miscellanea cassinese 39 (Montecassino: Badia di Montecassino, 1975), 117; and Louis Holtz, “Le Parisinus Latinus 7530, synthèse cassinienne des arts libéraux,” Studi medievali 16 (1975): 97–152. Similarly, book I was used as a source by Julian of Toledo and excerpts from it appear in a collection on the Liberal Arts with a Visigothic pedigree found in Erfurt, Biblioteca Amploniana, 2O 10 (9th c., in., western Germany); see Giliola Barbero, “Per lo studio delle fonti del Liber Glossarum: il MS. Amploniano F.10,” Aevum 67 (1993): 253–78.
 Above all, only Carolingian manuscripts of the separately transmitted first book of the Etymologiae contain glosses, which suggest a particularly active use in the school context. The glossing peaks at the end of the ninth century.
 The oldest manuscript in my corpus containing an excerpt collection is Paris, BnF, Lat. 12444 (8th/9th c., perhaps Fleury, prov.: Corbie) containing the fullest copy of the famous Collectio Sangermanensis. In its current form, this extensive collection including both excerpts from the Etymologiae, from other Isidore’s works, and from other texts, is presumably Carolingian, but it may be indebted to an older question-and-answer collection of Isidorian material.
 The oldest surviving manuscript containing an Isidorian excerpt collection is Paris, BnF, Lat. 12444 (8th/9th c., perhaps Fleury, prov.: Corbie). Most of excerpt collections were copied in the ninth century. The oldest surviving manuscripts containing the separately transmitted book I as an ars grammatica are three famous early Carolingian grammatical compendia: Berlin, Staatsbibliothek Preussischer Kulturbesitz, MS Diez. B. Sant. 66 (8th c., ex., northern Italy and northern France), Bern, Burgerbibliothek, MS 207 (8th/9th c., Fleury), and Vatican, Biblioteca Apostolica Vaticana, MS Pal. Lat. 1746 (8th/9th c. and 9th c., in., Lorsch)
 On these concerns and interests, see in particular Keefe, Water and the Word, 1:35–36.
 Compare with John J. Contreni, “The Carolingian Renaissance: Education and Literary Culture,” in The New Cambridge Medieval History, vol. 2 (Cambridge: Cambridge University Press, 1995), 726.
 While, it can be added, there are only a few grammatical compendia and pastoral collections from Italy, although, as was said, many Italian small Isidores perhaps remain to be identified.
 See footnote 73.
 The recasting of the Etymologiae into question-and-answer format has been treated as a symptom of an appropriation for school; Cardelle de Hartmann, “Uso y recepción de las Etymologiae de Isidoro,” 499–500. It should be, however, seen more broadly as a format used in a variety of educational context, and in the Carolingian period specifically in the training of the clergy; Keefe, Water and the Word, 1:147.
 The production of new redactions of the Etymologiae is briefly discussed in Reydellet, “La diffusion des Origines,” 388. Examples of additions can be found in Codoñer Merino, “Transmisión y recepción de las ‘Etimologías’,” 11–14.
 For example, it is the place of origin of several highly innovative manuscripts of the Etymologiae containing many additions and other novel features, such as Berlin, Staatsbibliothek, Hamilton 689 (11th c., perhaps northern Italy) and Berlin, Staatsbibliothek, Lat. fol. 641 (fols. 17-257) (9th c., med., northern Italy). It is also the home of the “contracted” family β of the Etymologiae, which is presumably an early redaction of Isidore’s encyclopedia. On this family, see Reydellet, “La diffusion des Origines,” 395–96 and 435–36.
 See Steinová, “Two Carolingian Redactions of the Etymologiae from St. Gallen.” Telling is also the absence of glosses to the first book of the Etymologiae in the two St. Gallen grammatical compendia containing this book, St. Gallen, Stiftsbibliothek, MS 876 (9th c., in.) and St. Gallen, Stiftsbibliothek, MS 882 (9th c., ¾), even though other texts in these manuscripts are rather heavily glossed.
 See Bischoff, “Die europäische Verbreitung der Werke Isidors von Sevilla,” 340. Moreover, one of the two St. Gallen redactions of the Etymologiae, which depends on a collation of several manuscripts, contains many more instances of collation of books XI-XX than books I-X.
 See Wallace Martin Lindsay, “New Evidence for the Text of Festus,” The Classical Quarterly 10 (1916): 106–15; and Claudia Villa, “Uno schedario di Paolo Diacono. Festo e Grauso di Ceneda,” Italia medioevale e umanistica 27 (1984): 56–80.
 These are Cesena, Biblioteca Malatesiana, S.XXI.5 (9th c., 1/3, northern Italy) and Ivrea, Biblioteca capitolare, MS LIII (37) (10th/11th c., Ivrea).
 This redaction can be easily identified because it incorporates Isidore’s De natura rerum as book IV into the Etymologiae and thus increases the total number of books to twenty-one. Its oldest witness is the mid-ninth-century manuscript now in Berlin mentioned in footnote 82.
 The best representative of this deviation from a regional trend is Laon, Bibliothèque municipale, MS 447 (9th c., 2/3, Mainz, prov.: Laon). This Big Isidore is the most heavily annotated manuscript of the Etymologiae from France featuring many clearly scholarly annotations rather than simply absorbing local school glosses. Authorities, especially Bede, are frequently copied in the margin to supplement Isidore. The most heavily annotated book is book XI on human body, while books I-III, which carry the most glosses in French manuscripts, have been barely touched by the same group of annotators from insular background. This manuscript is briefly mentioned in John J. Contreni, The Cathedral School of Laon from 850 to 930: Its Manuscripts and Masters, Münchener Beiträge Zur Mediävistik Und Renaissance-Forschung 29 (Munich: Arbeo-Gesellschaft, 1978), 45.
Andrés Sanz, María Adelaida. “Una reelaboración de textos isidorianos en forma de interrogationes et responsiones.” Helmantica 57, no. 172 (2006): 29–48.
Anspach, August Eduard. “Das Fortleben Isidors sim VII. bis IX. Jahrhundert.” In Miscellanea Isidoriana, 322-56. Rome: Universita Gregoriana, 1936.
Anspach, August Eduard. Taionis et Isidori Nova Fragmenta et Opera. Textos Latinos de La Edad Media Española 2. Madrid: Bermejo, 1930.
Barbero, Giliola. “Per lo studio delle fonti del Liber Glossarum: il MS. Amploniano F.10.” Aevum 67 (1993): 253–78.
Beeson, Charles Henry. Isidor-Studien, Quellen und Untersuchungen zur lateinischen Philologie des Mittelalters, 4.2. Munich: Beck, 1913.
Bischoff, Bernhard. Katalog der festländischen Handschriften des neunten Jahrhunderts: (mit Ausnahme der wisigotischen). Edited by Birgit Ebersperger. 4 volumes. Veröffentlichungen der Kommission für die Herausgabe der mittelalterlichen Bibliothekskataloge Deutschlands und der Schweiz. Wiesbaden: Harrassowitz, 1998-2017.
Bischoff, Bernhard. “Die europäische Verbreitung der Werke Isidors von Sevilla.” In Isidoriana: collección de estudios sobre Isidoro de Sevilla, edited by Manuel C. Díaz y Díaz, 317-44. León: Centro de estudios San Isidoro, 1961.
Bozzolo, Carla, and Ezio Ornato. Pour une histoire du livre manuscrit au Moyen Âge: Trois essays de codicologie quantitative. Équipe de recherche sur l’humanisme français. Textes et études 2. Paris: CNRS, 1980.
Cardelle de Hartmann, Carmen. “Uso y recepción de las Etymologiae de Isidoro.” In Wisigothica. After M. C. Díaz y Díaz, edited by Carmen Codoñer Merino and Paulo Farmhouse Alberto, 477-502. mediEVI 3. Florence: SISMEL, 2014.
Codoñer Merino, Carmen. “Transmisión y recepción de las ‘Etimologías’.” In Estudios de latín medieval hispánico. Actas del V Congreso Internacional de Latín Medieval Hispánico, edited by José Martínez Gázquez, Óscar de la Cruz Palma, and Cándida Ferrero Hernández, 5-26. Florence: SISMEL – Edizioni del Galluzzo, 2011.
Codoñer Merino, Carmen. Introducción al Libro X de las “Etymologiae”: su lugar dentro de esta obra, su valor como diccionario. Logroño: Logroño Fundación San Millán de la Cogolla, 2002.
Codoñer Merino, Carmen, José Carlos Martín, and María Adelaida Andrés Sanz, “Isidorus Hispalensis Ep.” In La Trasmissione Dei Testi Latini Del Medioevo/Medieval Texts and Their Transmission, edited by Paulo Chiesa and Lucia Castaldi, 274-417, vol. 2. Florence: SISMEL – Edizioni del Galluzzo, 2005.
Contreni, John J. “The Carolingian Renaissance: Education and Literary Culture.” In The New Cambridge Medieval History. Volume 2. Cambridge: Cambridge University Press, 1995.
Contreni, John J. The Cathedral School of Laon from 850 to 930: Its Manuscripts and Masters, Münchener Beiträge Zur Mediävistik Und Renaissance-Forschung 29. Munich: Arbeo-Gesellschaft, 1978.
Custodio Vega, Angel. “El Prof. Doctor August Eduard Anspach: semblanza literaria.” In Catálogo de los materiales codicológicos y bibliográficos del legado científico del Prof. Dr. August Eduard Anspach, edited by José María Fernández Catón, 9-28. León: Centro de estudios e investigación “San Isidoro,” 1966.
Díaz y Díaz, Manuel C. “Problemas de algunos manuscritos hispánicos de las Etimologías de Isidoro de Sevilla.” In Festschrift Bernhard Bischoff zu seinem 65. Geburtstag, edited by Johanne Autenrieth and Franz Brunhölzl, 70-80. Stuttgart: Hiersemann, 1971.
Díaz y Díaz, Manuel C. Los capítulos sobre los metales en las “Etimologías” de Isidoro de Sevilla. León: Cátedra de San Isidoro, 1970.
Elfassi, Jacques. “Isidore of Seville and the Etymologies.” in A Companion to Isidore of Seville, edited by Andrew Fear and Jamie Wood, 245-78. Boston and Leiden: Brill, 2020.
Elfassi, Jacques. “Nuevas fuentes en la biblioteca de Isidoro de Sevilla.” In Latinidad Medieval Hispánica: Congreso Internacional de Latín Medieval Hispánico, 2013, La Nucia, Spain, edited by Juan Francisco Mesa Sanz, 107-16. MediEVI 14. Florence: SISMEL – Edizioni del Galluzzo, 2017.
Elfassi, Jacques. “Connaître la bibliothèque pour connaître les sources: Isidore de Séville.” Antiquité tardive 23 (2015): 59–66.
Fernández Catón, José María. Catálogo de los materiales codicológicos y bibliográficos del legado científico del Prof. Dr. August Eduard Anspach. León: Centro de estudios e investigación “San Isidoro,” 1966.
Fernández Catón, José María. Las “Etimologías” en la tradición manuscrita medieval estudiada por el Prof. Dr. Anspach (León: Centro de estudios e investigación “San Isidoro,” 1966.
Ferraces Rodríguez, Arsenio, ed. “Isidorus medicus”. Isidoro de Sevilla y los textos de medicina. Monografias 113. Coruña: Universidade da Coruña, 2005.
Fontaine, Jacques. Isidore de Séville et la culture classique dans l’Espagne wisigothique. 2nd edition Paris: Études augustiniennes, 1983.
Fontaine, Jacques, et al. “Compte rendu du colloque isidorien tenu à l’Institut d’études latines de l’Université de Paris le 23 juin 1970.” Revue d’Histoire des Textes 2, no. 1972 (1973): 282–88.
Fontaine, Jacques. “La diffusion de l’oeuvre d’Isidore de Séville dans les scriptoria helvétiques du haut Moyen Âge.” Schweizerische Zeitschrift für Geschichte 12 (1962): 305–22.
Forstner, Karl. “Ergänzungen zu B. Bischoffs Hss Katalog (Salzburger Fragmente).” Scriptorium 62 (2008): 122–38.
Ganz, David. “Book Production in the Carolingian Empire and the Spread of Caroline Minuscule.” In The New Cambridge Medieval History 2: C. 700-c. 900, edited by Rosamond McKitterick, 786-87. Cambridge: Cambridge University Press, 1995.
Gorman, Michael. “The Carolingian Miscellany of Exegetical Texts in Albi 39 and Paris Lat. 2175.” Scriptorium 51 (1997): 336–54.
Greenia, George D. “The Bigger the Book: On Oversize Medieval Manuscripts.” Revue Belge de Philologie et d’histoire 83 (2005): 723–46.
Havel, Dalibor. Počátky latinské písemné kultury v českých zemích: nejstarší latinské rukopisy a zlomky v Čechách a na Morave. Opera Facultatis Philosophicae Universitatis Masarykianae 479. Brno: MUNI, 2018.
Holtz, Louis. Donat et la tradition de l’enseignement grammatical: étude sur l’Ars Donati et sa diffusion (IV. – IX. siècle) et éd. crit. Paris: CNRS, 1981.
Holtz, Louis. “Le Parisinus Latinus 7530, synthèse cassinienne des arts libéraux.” Studi medievali 16 (1975): 97–152.
Keefe, Susan. Water and the Word: Baptism and the Education of the Clergy in the Carolingian Empire. Volume 1. Publications in Mediaeval Studies. Notre Dame: University of Notre Dame Press, 2002.
Kurrus, Theodor. “In memoriam: August Eduard Anspach (1860-1943).” In Spanische Forschungen der Görres-Gesellschaft 1, edited by Johannes Vincke, 349-56. Gesammelte Aufsätze zur Kulturgeschichte Spaniens 26. Münster: Aschendorff, 1971.
Lapidge, Michael. “An Isidorian Epitome from Early Anglo-Saxon England.” Romanobarbarica 10 (89 1988): 443–83.
Law, Vivien. The Insular Latin Grammarians. Studies in Celtic History 3. Woodbridge: Boydell Press, 1987.
Lindsay, Wallace Martin. “New Evidence for the Text of Festus.” The Classical Quarterly 10 (1916): 106–15.
Lindsay, Wallace Martin. Etymologiarum sive Originum libri XX. 2 volumes. Oxford: Clarendon Press, 1911.
Lowe, Elias Avery. Codices Latini Antiquiores: A Palaeographical Guide to Latin Manuscripts Prior to the Ninth Century. 11 volumes. Oxford: Clarendon Press, 1934-66.
Maniaci, Marilena. “Costruzione e gestione dello spazio scritto fra Oriente e Occidente: principi generali e soluzioni specifiche.” In Scrivere e leggere nell’alto Medioevo, 473-512. Settimane di studio del Centro italiano di studi sull’alto Medioevo 59. Spoleto: Centro Italiano di Studi sull’Alto Medioevo, 2012.
Manitius, Max. Geschichte der lateinischen Literatur des Mittelalters. Volume 1. Munich: Beck, 1911.
Maassen, Friedrich. Geschichte der Quellen und der Literatur des Canonischen Rechts im Abendlande bis zum Ausgange des Mittelalters. Volume 1: Die Rechtssammlungen bis zur Mitte des 9. Jahrhunderts. Graz: Leuschner & Lubensky, 1870.
McKitterick, Rosamond. The Carolingians and the Written Word. Cambridge: Cambridge University Press, 1989.
Parkes, Malcolm B. Pause and Effect: An Introduction to the History of Punctuation in the West. Aldershot: Scolar, 1992.
Pellegrin, Elisabeth. “Fragments et membra disiecta.” Codicologica 3 (1980): 70-95.
Porzig, Walter. “Die Rezensionen der Etymologiae des Isidorus von Sevilla. Vorbemerkung.” Hermes 72, no. 2 (1937): 129–70.
Reydellet, Marc. “La diffusion des Origines d’Isidore de Séville au Haut Moyen âge.” Mélanges d’archéologie et d’histoire 78, no. 2 (1966): 383–437.
Schindel, Ulrich. “Zur frühen Überlieferungsgeschichte der Etymologiae Isidors von Sevilla.” Studi medievali 29, no. 2 (1988): 587–605.
Schütz, Heinz-Albert. “Die Schrift ‘De medicina’ des Isidor von Sevilla: ein Beitrag zur Medizin im spätantiken Spanien.” Ph.D. diss., University of Giessen, 1984.
Stadelmaier, Michael. Die Collectio Sangermanensis XXI titulorum: eine systematische Kanonessammlung der frühen Karolingerzeit. Freiburger Beiträge zur mittelalterlichen Geschichte 16. Frankfurt am Main: Peter Lang, 2004.
Steinová, Evina. “Two Carolingian Redactions of the Etymologiae from St. Gallen.” Mittellateinisches Jahrbuch, forthcoming.
Steinová, Evina. “The Materiality of Innovation: Formats and Dimensions of the Etymologiae of Isidore of Seville in the Early Middle Ages.” In Entangled Manuscripts: 600-1200, edited by Anna Dorofeeva and Michael J. Kelly (Binghamton: Gracchi Books, forthcoming 2021).
Tardif, Ernest-Joseph. “Un abrégé juridique des Étymologies d’Isidore de Séville.” In Mélanges Julien Havet, 659-81. Paris: Ernest Leroux, 1895.
Warntjes, Immo. “Isidore of Seville and the Formation of Medieval Computus.” In A Companion to Isidore of Seville, edited by Andy Fear and Jamie Wood, 457-523. Boston and Leiden: Brill, 2020.
Van den Abeele, Baudouin. “La tradition manuscrite des Étymologies d’Isidore de Séville.” Cahiers de recherches médiévales et humanistes 16 (2008): 195–205.
Van de Vrie, Esther. “Een tekst voor studie, toen en nu.” In Perkament in stukken: Teruggewonden middeleeuwse handschriftfragmenten, edited by Bart Jaski, Marco Mostert, and Kaj van Vliet, 168-71. Hilversum: Verloren, 2018.
Van Els, Ad. “Een leeuw van een handschrift: Ademar van Chabannes en MS Leiden, Universiteitsbibliotheek, Vossianus Latinus Octavo 15.” Ph.D. diss., Utrecht University, 2015.
Villa, Claudia. “Uno schedario di Paolo Diacono. Festo e Grauso di Ceneda.” Italia medioevale e umanistica 27 (1984): 56–80.