The first two chapters in this volume bring to life Augustine as a preacher in the context in which he practiced the craft of preaching so expertly and in his goal of receiving God’s love in grace. They tell us of the environment in which the sermons came into being and of the conditions and intrinsic motivations that enabled Augustine to become an important – if not the most important – preacher of his time. The chapter you are currently reading takes a different part of the history of Augustine’s sermons as its focal point. Rather than the context of their genesis, this chapter will focus on the afterlife of the sermons, their transmission. It will encompass the many centuries from Late Antiquity, across the Middle Ages, up to today, that the sermons traversed in order to be available to us. Most importantly, it will focus on how this long journey affected the individual sermons, the corpus as a whole, and the evolving perception of Augustine as preacher and Church Father. As such, it will place some necessary caveats to how Augustine’s sermons can be used and studied as sources to reconstruct the context of early Christianity.
The exposition on this vast stretch of time will be necessarily brief; the main goal of the chapter will be to explain in what forms the corpus of Augustine’s preaching we identify as Sermones ad populum was transmitted in the Middle Ages and how that transmission affected the texts as we read them today. This explanation begins with an uncomfortable truth. When using a modern critical edition of Augustine’s sermons, the reader may be forgiven for presuming that if the editor did their job correctly, the printed text would reconstruct with a high level of accuracy the exact words as preached by Augustine. This is simply not the case. I am not at all disparaging the work of the editors, but rather pointing out that any critical reconstruction of the text of a sermon is based on the surviving “witnesses” to this text: medieval manuscripts. For the corpus of Augustine’s preaching, these medieval manuscripts most often do not predate the 9th century, some 500 years after Augustine spoke the words. Inevitably this distance has impacted both the text, which needed to be copied by hand, and the corpus as a whole, with many texts irretrievably lost. The medieval reception of Augustine’s sermons has meant that only rarely are we able to reconstruct Augustine’s words with confidence. Very often, we know that what we have is an abbreviated and altered version of what Augustine actually preached. So, how do we determine which sermons can bring us back all the way to Late Antiquity, and which cannot? And how do we use the latter responsibly as sources?
The limits on our access to Augustine’s context and actual words are greater for his homiletic corpus than for other records of his output because of certain characteristics of sermons, which make them particularly vulnerable. To round out this introduction, I will briefly describe these characteristics.
First, sermons are relatively short texts, so in the context of a physical carrier – a scroll or a codex – they tend to be bundled together, forming clusters and collections which increased in size over time. Second, because they were short, individual sermons were rarely known by a generally recognized title, so they were difficult to identify outside of the context of an established collection. This also makes them vulnerable to misattribution. Third, Augustine famously did not systematically organize and review his homiletic corpus. While the Enarrationes in psalmos, the 124 In Iohannis euangelium tractatus, and the In Epistolam Iohannis ad Parthos Tractatus Decem are commentaries comprised of homiletic material and thus formed relatively stable collections, what we know under the umbrella term of Sermones ad populum traversed the Middle Ages in an ever-shifting assortment of smaller and larger collections. Finally, Augustine’s sermons, like those of other patristic preachers, were actively used as part of medieval religious life. This was not a static, archival tradition, but a corpus that served several purposes throughout the Middle Ages and was adapted to fit its new contexts and uses. With these characteristics as a guide, let us now explore the late-antique and medieval transmission of Augustine’s sermons.
The Earliest Dissemination of Augustine’s Sermons
By now it is well known that Augustine improvised the majority of his sermons, reflecting on the scriptural reading of the day, often incorporating long digressions, and adapting in the moment to reactions of the audience. Professional stenographers, who might be close associates of the Bishop himself but who might also be employed by private audience members or commercial agents, recorded his words.Footnote 1 The library at Hippo was an important repository for these records, but there is also evidence of sermon records created and collected on the occasions of Augustine’s visits to Carthage, one of several places outside Hippo where he preached.Footnote 2
The two most important contemporary records that inform us on Augustine’s preaching activity are his own comments in the Retractationes and the Vita Augustini,Footnote 3 a concise biography written between 431 and 439, shortly after Augustine’s death, by Possidius of Calama, Augustine’s friend and disciple.
From the Retractationes we learn that, while Augustine had the intention to edit his homiletic corpus, he never got around to it. Unlike for the Enarrationes and the two sets of Tractatus, Augustine, to our knowledge, never created clearly defined and identifiable ensembles for the corpus of sermons which is now termed the Sermones ad populum. There is an upside to this circumstance. Having never been edited by their author, the majority of Augustine’s sermons started to circulate in their “raw” state, and originally retained more oral features than they would have if Augustine had polished them. However, these features were subject to a process of erosion over the course of the sermons’ medieval transmission. Historical references that had lost their meaning to medieval audiences, long digressions that veered too far off topic, ad hoc reactions to the audience or the circumstances as Augustine was preaching – these elements tended to disappear as the sermon was copied from manuscript to manuscript, collection to collection.
Possidius paints a vivid picture of Augustine’s popularity as a preacher (e.g., Vita 31.9), but more importantly for our purpose, he annexes to his biography a bibliography of Augustine’s literary production, including his sermons. The bibliography is organized according to the main adversaries Augustine faced during his episcopal career, with lists of sermons per heresy. At the end of the bibliography (X/6) is a list of some 200 entries describing sermones uarii. It is generally agreed that this list far from covers the entirety of Augustine’s homiletic production, but we cannot be certain what percentage of the total it represents. Possidius’ bibliography reveals some interesting elements about the first stage of transmission of Augustine’s sermons.
The sermons are often described in groups, comprising sometimes two or three but also larger clusters. Clustering of sermons thus seems to have existed from the beginning. The titles Possidius gives to the texts vary from short and generic (“Love, two sermons”) to relatively detailed descriptions of the scriptural passages referenced in the text.
If Possidius’ list reflects the organization of the library at Hippo – which is not certain, but seems likely because series from this list can be detected in some of the oldest sermon collections – then the original organizing principle for the sermons seems to have been based on subjects rather than, for example, chronology or location.
Despite these valuable resources, the earliest transmission of Augustine’s preaching – how, where and why his sermons were disseminated – remains shrouded in mystery.
Late-Antique and Early Medieval Augustinian Sermon Collections
The primary mode of transmission for Augustinian sermons is in the form of collections. These collections were subject to significant reorganization over the course of the Middle Ages, with new collections building on older ones. This means that a single sermon can be found in multiple collections. This is a context in which the restrictions of the physical process of manuscript transmission engage with the creative intentions of medieval compilers.
The term “sermon collection” is not particularly easy to define. Technically, any combination of sermons found within the confines of a codex could be termed a collection. It is, after all, a deliberate selection of materials brought together as one whole. However, in the context of patristic sermon studies, the term “collection” is defined more narrowly and generally used to denote the following types of compilation:
a group of sermons that can be shown to have already existed as such in Late Antiquity (even if no physical witnesses from that period remains);
a group of sermons which exhibit a clear reason for traveling together, for example, because the texts share a thematic connection;
a group of sermons that was perceived by medieval readers and copyists as a (stable) collection, usually has a stable title, and was copied with minimal variation in multiple manuscript witnesses.
This definition is not perfect. In the context of Augustine’s preaching, groups of sermons tend to be termed “collections” based on scholarly precedents that have become widely accepted and newly identified collections have difficulty in gaining traction in the scholarly community. The division between manuscripts that contain recognized collections and the so-called eclectic manuscripts (manuscripts that contain combinations of sermons that do not fall in any of the categories delineated here) is really an artificial one. This section will present the collections that have a connection to the late-antique and early medieval (i.e., pre- or early Carolingian) period, while the next section will focus on collections put together in later centuries. This distinction is important, because the type of collection in which a sermon travels can help determine how the quality of its text should be evaluated.
There are certain features that unite the oldest collections of Augustine’s sermons. They are generally single-author collections, meaning they contain only material preached by Augustine, though sometimes sermons wrongly attributed to him have attached themselves to the main corpus of authentic Augustinian sermons. They are usually relatively small collections, between around twenty and fifty sermons. And their dissemination in their “pure” form was almost always very limited. Most of the collections which have been identified as stemming from Antiquity are preserved in no more than a handful of manuscript witnesses, which usually date to much later. A broader dissemination of part of such a collection does occur when it is taken up in a later, newly formed medieval compilation with a much wider reach. The organizing principle for these collections is often spatio-temporal. One finds clusters of sermons preached in Carthage or clusters preached consecutively during a particular part of the year, an Easter cycle, for example. An important method of identifying antique collections consists of matching a particular series of sermons in the collection to the same series in Possidius’ bibliography.
The most important and up-to-date overview on the transmission of Augustine’s Sermones ad populum is François Dolbeau’s magisterial entry in the Augustinus-Lexikon (Dolbeau Reference Dolbeau2019–2021), which identifies at least ten collections of Sermones ad populum with a clear connection to Late Antiquity.
Mainz-Grande Chartreuse – K. By far the most famous of the Augustinian sermon collections today is the Mainz-Grande Chartreuse collection. Discovered by François Dolbeau in 1990Footnote 4 in the codex Mainz Stadtbibl. I 9 (15th century), the collection of at least thirty sermons was present in the library at Hippo, as recorded by Possidius in Indic. X6 101–135. Two hypotheses have been put forward with regard to the origin and internal order of the collection. Either it represents a series from the start of Augustine’s episcopal career – perhaps the year 397, in which a few other clusters have been introduced – or the collection is comprised of small blocs preached throughout his career and put together later as a liturgical cycle.
Sessoriana – S. This collection, named after the Roman library which once held its principal witness (Roma Bibl. Naz. Vitt. Em. 1357, 8th–9th century), knew a limited circulation in Italy, with four extant witnesses. It consists of twenty-two items, including, in addition to Augustinian sermons, also a treatise, a sermon now deemed inauthentic (s. 391), and three extracts from De diu. quaest. Some smaller series from September and October 417 seem to have retained their original order of preaching. Particularly interesting in this collection are the titles, which refer to specific dates and also to nine different basilicas in Carthage, without ever specifying the name of the city, indicating that it was obvious to the compiler where these basilicas were located. This observation, plus the fact that very few of the Sessoriana sermons feature in the Indiculum, have led to the hypothesis that this collection was the work of a notary or bookseller from Carthage.
Collectio Campana – N. As the name suggests, this collection’s two extant witnesses, Monte Cassino 17 and Napels Bibl. Naz. Vienn. lat. 14, were written in Campania, both in the 11th century. The collection is comprised of two African series of sermons, with some material – extracts from en. Ps., as well as a few inauthentic sermons – added to them. The first series contains twenty-two texts preached in Hippo, the other holds twenty-one texts preached in Carthage. The second series is accompanied by indications of place and date of preaching and by a reference to “Aurelius A.” as the sermons’ author in the Montecassino manuscript. In the Napels codex, the collection is enriched by nine sermons borrowed from the Collectio Sessoriana.
Mainz-Lorsch – Z. The second half of the famous Mainz codex, this collection was, prior to Dolbeau’s discovery, known only through inventories of the library of Lorsch, which described a collection of twenty-nine titles. The Lorsch description seems to reflect a more original state of the collection, including in particular the mention of rare place-names located in the African hinterland. The majority of the sermons in this collection can also be found in Possidius’ Indiculum. This fact, combined with the mention of pagans, Donatists and Arians in the titles has led Dolbeau to conclude that this is a very old collection of polemical sermons.
De alleluia – A. This is a collection of forty-seven authentic Augustinian sermons, extant in five manuscripts from northern France (a sixth witness contains only the capitula of the collection: a table of contents without the actual texts), of which the oldest dates to the 9th century. All six witnesses present a disordered and mutilated version of the collection. The reconstructed collection combines three individual late-antique series. The first (items 1–9) represents a series of sermons preached in Carthage in May–June 418. The second (items 10–19) forms an Easter sequence preached in Hippo, perhaps from a single year. The third is a winter series preached after 425 (which we know due to reference to the miracles of St. Stephen), formed by items 25–33, while items 20–24 and 34–47 constitute the “leftovers.” None of the three series correspond to any item in the Indiculum. This is a precious collection, because it transmits many sermons not extant anywhere else, but also because it provides rare clues to the earliest modes of transmission of Augustine’s sermons.
De paenitentia – P. This collection exists in only three manuscripts, one of which, the model for the other two, is a codex made of papyrus dating to around 700,Footnote 5 unfortunately very damaged. The direct witnesses are all from eastern France. The collection itself comprises thirty-seven sermons, but the first and last items may have been added at a later stage to an older nucleus of twenty-eight pieces (one among them a post sermonem). According to Dolbeau Reference Dolbeau2019–2021, nine sermons were preached in Carthage in June 410 and over the course of 411. In its extant witnesses the collection is preceded by a collection of sixteen Augustinian letters. While the actual antique origins of this collection as a whole are in doubt, the precious papyrus codex can tell us about medieval uses of these older collections; it was in the possession of Florus of Lyon in the 9th century and bears annotations made by him.
Bobbio – B. A single witness, Vat. Lat. 5758, copied shortly before 640, testifies to three series of sermons, which are all attributed to Augustine, though the first, twenty-one items, is actually authored by Peter Chrysologus. The second series of twenty-four texts and the third series of five texts, separated by a sermon incorrectly attributed to Augustine, round out the collection. These thirty pieces contain in total twenty-two Augustinian sermons, of which twelve were identified also in the Indiculum. Overall, the textual quality varies, and, as Dolbeau Reference Dolbeau2019–2021 points out, especially for those sermons also witnessed in the Mainz-Lorsch collection, the quality of the text transmitted by the Bobbio manuscripts is inferior. The collection was drawn on by two medieval homiliaries, Paris BnF lat. 792 (11th century) and Piacenza Bibl. Cap. 60–63 (12th century).
De bono coniugali – C. This collection consists of a combination of fourteen sermons, one of which is inauthentic, five letters, and seven treatises, of which five are Augustinian and two ascribed to Nicetas of Remesiana. Among the thirteen authenticated Augustinian sermons are two series, one containing three sermons and the other four, which coincide with series in Possidius’ Indiculum. The first ten items of the collection seem directed toward the moral education of different stages of life. All four extant witnesses of De bono coniugali have as their ancestor the slightly mutilated manuscript Vat. Pal. lat. 210, which was written in Italy at the beginning of the 6th century. While the collection used to be considered Arlesian – compiled and potentially rewritten by Caesarius of Arles (ca. 470–542), as discussed in the next section of this chapter – recent researchFootnote 6 has identified in Vat. Pal. Lat. 210 annotations which can be attributed to Eugippius (ca. 460 – ca. 535), which makes it impossible that Caesarius was responsible for this collection.
Cluny – Y. This is, again, a collection preserved in a single witness, Brussels Bibl. Roy. 14920–14922 (11th century). This manuscript contains a sizable compilation, the first part of which is comprised of thirty-seven sermons borrowed from De paenitentia, De bono coniugali, and De alleluia. After this follows an antique series of fifteen sermons (one of which is incorrectly attributed to Augustine) preached in Carthage and Hippo, including five sermons covering the Monday through Friday of a single Octave of Easter. It is likely that this series is a selection from an otherwise lost antique collection.
Wolfenbüttel – W. The final collection in this list is also the largest, with ninety-five to ninety-seven texts, of which over seventy are authentic Augustinian sermons, many not found elsewhere. The collection is organized as a liturgical cycle. The collection has gaps at the beginning, and the opening series contains fifteen of the twenty-two non-Augustinian pieces present in the collection, attesting to an addition to the Augustinian nucleus. Overall, this collection retains fewer features typical of antique collections, with generic titles and a reshuffling of the sources. The assumption of its antique character is, therefore, derived mainly from the large number of rare sermons, as well as the fact that for some of its sermons, it preserves a longer or interpolated version. This collection is, once again, preserved in a single manuscript, Wolfenbüttel Herz.-Aug. Bibl. cod. 4096 (Guelf. 12 Weiss.), created around 900 in northern Germany.
Two observations with regard to this overview are particularly noteworthy. First, these collections show variety in the original or early ways of organizing and circulating Augustine’s sermons. There are clusters that reflect chronologically and spatially connected sermons. There are some thematic series, but also clusters centered on a single feast. Most of these collections are strictly single-author, but De bono coniugali, with one of the oldest extant manuscript witnesses for Augustine’s preaching, shows that there are exceptions. The collections consist of twenty sermons or more, but in several of them, smaller series can be isolated. The exact time at which these smaller series were combined into larger collections is difficult to pinpoint.
Second, there are examples in these collections of sermons which are still adorned with markers of their antique context – the names of the basilicas where they were preached in the title, the phrase conuersi ad dominum at the end – and yet the text accompanying these signs of the sermon’s original context is not necessarily the most authentic or “best” text. There is nothing preventing a compiler from keeping these markers but still – accidentally or deliberately – altering the text. An old collection does not necessarily – or at least not always – equal the most authentic or pure text. Frequently a cobbling together of textual features from different branches of the tradition is required. This makes the correct use of the sermons as sources very complicated.
From Antiquity to the Middle Ages: Caesarius of Arles
Since the first overview of Augustinian sermon collections, created by Cyril Lambot as part of the introduction to his critical edition of Sermones ad populum 1–50 (CCSL 41), a separate category has been devoted to so-called Arlesian collections: collections that bear the mark of Caesarius of Arles. He was an avid preacher and advocated strongly – and ultimately successfully – for expanding the right to preach to priests and increasing the congregation’s exposure to sermons.Footnote 7 To supply the priests in his diocese with inspiring and doctrinally sound preaching materials, he put together a number of collections, incorporating a lot of Augustinian material. Importantly, he freely adapted the source material, adding or supplanting paragraphs, cutting out digressions and generally refitting the sermons to better suit his purpose. Using the intensity of Caesarius’ interventions as a guide, Germain Morin has divided the corpus into Caesarian sermons with an Augustinian core and Augustinian sermons with a Caesarian adaptation.Footnote 8 We can recognize the alterations Caesarius made when a sermon he included is preserved in its more original state in a parallel collection, but in those case where Arlesian collections are our only extant witnesses, we must operate in the knowledge that it is highly likely that the text of the sermon was manipulated.
There are three extant collections currently considered to be the work of Caesarius, but the marks of his intervention are not always unambiguous. The most widely diffused of the Arlesian collections, with over sixty extant witnesses, is the Quinquaginta homiliae, which contains fifty sermons, of which about half are considered to retain enough of their original state to be identified as Augustinian. The Collectio Lugdunensis is preserved in a single, incomplete witness from the 7th century, currently dispersed over three manuscripts (Lyon Bibl. Mun. 788 + 604 + Paris Bibl. Nat. n.a.l. 1594). The collection as we know it now consists of some twenty-three pieces, of which a handful are not authentic. Finally, the Marmoutier collection, comprised of a series of Augustinian sermons followed by texts of Maximus of Turin, Zeno of Verona, and Basil the Great, is preserved in a single manuscript, Tours Bibl. Mun. 279 (9th century), alongside works of Jerome. Several of the Augustinian sermons it contains are also found in the Collectio Lugdunensis. Caesarius, desiring to give the clergy of southern Gaul the tools to preach well, highlights what will be the quintessential medieval attitude to the Sermones ad populum. It was considered a corpus of significant interest, but first and foremost for its applicability to the medieval context, and it was adapted accordingly.
The Medieval Reception of Augustine’s Sermons
While Augustine’s sermons were certainly infused with his authority in the eyes of medieval compilers and readers, they were also viewed as very malleable objects in the Middle Ages, susceptible to extensive rewriting depending on the context in which they were reused. The previous section of this chapter dealt with sources in which traces of the late-antique context are still discernible, but these sources represent only a small part of the medieval transmission of Augustine’s preaching. The vast majority of extant manuscript witnesses contain sermon collections that did not exist as such in Late Antiquity but were formed based on older sources, some of which are extant but many of which are lost to us today. These collections are of various types, created or further developed to accommodate specific needs.Footnote 9
The most widespread medieval collection of Augustine’s sermons, De uerbis Domini et Apostoli, brings together over sixty-four homilies on the Gospels and thirty-four texts on the New Testament Letters to form an exegetical sequence. The collection, extant in hundreds of witnesses, may have originated in northern Italy, in the early 8th century at the latest, as it was almost certainly mentioned in a catalogue of Fontenelle, Normandy which dates to 745. It is partly based on the Collectio Sessoriana, and must also have had access to now lost antique collections or libelli, from which it drew some forty pieces. It was reviewed and expanded in later times, again on the basis of rare sources.Footnote 10
In several of these medieval collections the organizational principles are not obvious to us. Some collections have simply pasted together two or more older models, with minimal alterations to the sequence of texts. Examples of this are the De lapsu mundi (L) collection and the Collectio Bruxellensis (X). The former (L) is a collection of twenty-five sermons borrowed from the Lyon and De alleluia collections and extant in some twenty manuscripts, none predating the 12th century. The latter (X) is the larger collection in which the Cluny series is also found. The manuscript Brussels Bibl. Roy. 14920–14922 contains selections from several older collections (P, C, A), largely preserving the order of its sources. Collections such as these may have functioned in the context of private reading, as stipulated in the rule of St. Benedict (d. ca. 550), or as model sermons, though other uses are also possible.
Often medieval collections mix sermons from different authors. Still, it can be possible to recognize antique series or small clusters that were at the basis of these collections. For example, a multi-author compilation preserved in Paris Bnf, Lat. 1771 from the first half of the 9th century in Fulda, Germany, contains a series of ten sermons attributed to Augustine – six are authentic – which reflect antique origins in the rarity of the pieces, antique features in the titles, and a correspondence to a sequence in Possidius’ Indiculum. Such series in medieval collections can even add previously unknown sermons to our corpus. A well-known example of this is the Erfurt collection, extant in a unique witness, Erfurt Univ. Bibl. Dep. Erf. CA. 12°11, 12th century, possibly of English origin.Footnote 11 A series of twenty-eight. pieces, mainly attributed to Augustine, contains nineteen authentic Augustinian sermons, of which six were (partially) unknown prior to the manuscript’s discovery. Medieval alterations to the text are in evidence, but so are connections to the Indiculum.
By far the most popular organizational principle for patristic sermons in medieval contexts was the homiliary, a collection which follows the cycle of the liturgical year, offering one or more sermons for each occasion, a feast or saint’s day, in the calendar. These liturgical collections existed already very early on, but they grew throughout the early Middle Ages to gargantuan proportions, with multi-author collections of 200–300 sermons in evidence by the 8th century.Footnote 12 Several 8th- and 9th-century homiliaries, such as those preserved in Orléans Bibl. Mun. 155, Vienna Österr. Nationalbibl. lat. 1616 and Vat. lat. 3828, seem to have an African basis. In the homiliaries we find sermons that are part of the corpus of Sermones ad populum, but equally the Tractatus in Iohannem and – to a lesser extent – the Enarrationes in Psalmos are mined for relevant materials.
The two most influential homiliaries are the so-called Roman homiliary, which exists in different incarnations, and the homiliary of Paul the Deacon. The former was relatively rich in Augustinian material, with twenty-seven authentic sermons and forty-eight items erroneously attributed to Augustine in the version of Alanus of Farfa, while Paul the Deacon included just three Sermones ad populum (there are more borrowings from other Augustinian works). Frequently, homiliaries built on one or both of these two “types” through customization. The “Roman homiliary” is a construct based on several extant sources: the homiliary of Agimundus preserved in Vat. lat. 3835/6 (8th century), a later version of the St. Peter homiliary in Vat. Arch. S. Pietro C 105 (9th–10th century), the collection of Egino of Verona in Berlin Staatsbibl. Phillips 1676 (8th century), and the homiliary of Alanus of Farfa, extant in multiple manuscripts and the incarnation of the Roman homiliary that was most commonly used for later customization. An example of a particularly Augustine-rich homiliary is that of Jouffroy (Vat. lat. 3828), in which almost half of the ninety-eight pieces are authentic Augustinian sermons.
While the sermons in homiliaries are typically highly developed, very often abbreviated, and made more generic to align them to the liturgical context, and while large numbers of wrongly attributed sermons have been introduced, these liturgical compilations can still contain valuable traces of their sources. Two examples of these are the Collectio Colbertina and the Sancti Catholici Patres. The former takes its name from the 12th-century manuscript that is its sole witness, Paris BnF Lat. 3798. It contains fifty-four pieces of which only a few are inauthentic; fifty-two of these also appear in the Sancti Catholici Patres, a massive collection of 345 sermons destined for reading in the refectory. This is a mixed-author collection, but in which the dominant presence of authentic Augustinian pieces testifies to a thorough research of available Augustinian collections, at least one of which is lost today. The Sancti Catholici Patres, often divided into two volumes, are extant in some fifteen manuscripts from the 12th century onwards.Footnote 13
The Sancti Catholici Patres already evinced a desire to put together a corpus of Augustinian sermons that was complementary to an existing collection – in this case De uerbis Domini et Apostoli. In the later Middle Ages, we also find collections that have the ambition to gather large numbers of Augustinian sermons. An example is the Collectio de diuersis rebus, preserved in a unique witness (though there were once more) from 12th-century Clairvaux, Troyes Bibl. Mun. 40.10. The collection of fifty texts is divided into a series on the Old Testament and a series on Christian virtues. Much large collections of this type, which apparently strove to compile a sort of opera omnia of Augustine’s preaching, are also in existence. The Collectio tripartita (probably 13th century) attributes all of its 263 articles to Augustine, sometimes wrongly. The Collectorium of Roberto de’ Bardi (chancellor at the Sorbonne from 1336 to 1349) gathers almost 600 homilies, but many of them are inauthentic.
Overall, the medieval transmission of Augustine’s sermons shows an ever-increasing distance from the original shape of the texts, titles, and attributions. However, the collections are often important and impactful in the medieval cultural and religious context, and in several cases they also have text-critical value, allowing us to reconstruct antique collections or access parallel transmissions.
Alternative Modes of Transmission
While collections of various types were the most common mode of transmission for Augustine’s sermons in the Middle Ages, there are also examples of sermons transmitted in isolation, meaning they are not consistently accompanied by other sermons but instead combined with other genres of texts, such as treatises or letters. The shift of a sermon’s transmission from collection to another context, or vice versa, could happen at any time in a sermon’s reception. However, there are only few sermons that appear to have circulated from the very beginning outside of the context of sermon collections. In fact, this means that, whatever their origins, they were almost immediately characterized or read as treatises rather than as sermons (e.g., s. 9 (De decem chordis)). Particularly famous sermons, such as s. 355–356 (De moribus clericorum), circulated in isolation throughout the Middle Ages (though we also find them as part of collections). Augustinian sermons are found excerpted in anthologies and compilation commentaries and entered into canon-law collections. There are also examples of Augustinian sermons that were transmitted among other works of Augustine, such as his letters, as well as enarrationes or tractatus that have been erroneously considered part of the sermones ad populum and vice versa.
Medieval Reuse of Augustine’s Preaching and its Impact on the Text and Corpus
An overview of the medieval transmission of Augustine’s preaching would not be complete without a section devoted to the vast corpus of dubiously attributed Augustinian preaching. The corpus of misattributed sermons can be divided into three categories:
sermons that have been authenticated by modern scholarship as being by Augustine but have been attributed in manuscripts to other authors, which happened throughout the Middle Ages;
sermons whose author (either late-antique or medieval) is unknown, but which have been wrongly attributed to known authorities in the manuscripts;
sermons that are reworkings, alternative versions, or patchworks, in which pieces of the original, authenticated or anonymous text can still be discovered.
Already very early on, sermons which were not actually his circulated under the name of Augustine. Over the course of the Middle Ages, Augustine remains the most used incorrect misattribution, with texts seemingly drawn to his name like moths to a flame. Augustine’s authority as a Father of the Church is, of course, part of the reason, but the dearth of authorized and clearly identified collections within his corpus of Sermones ad populum probably contributed as well. It was very difficult to verify the author of a sermon outside of such collections.
Scholarly treatment of the corpus of misattributed Augustinian preaching has mostly centered around authenticity critique, categorizing sermons as authentic or inauthentic. The category of dubii, meaning sermons of doubtful attribution, was an important part of the edition created by the Maurist editors in the late 17th century. The category gradually came into disuse over the centuries, although pleas are heard now to reinstate it as a tangible witness to the difficulties we have in securely identifying the author of a sermon and recognizing the medieval interventions to a text. There has been much discussion on the criteria by which an authentic Augustinian sermon can be recognized.Footnote 14
In fact, the rigid distinction between authenticated and inauthentic sermons obscures the reality that many of the sermons currently considered authentic are medieval versions – reworked and abbreviated – of the Augustinian sermons as they were originally preached. The discovery of the Dolbeau sermons was instrumental to this conclusion. The Mainz codex gave us many long sermons in a shape which we can recognize as close to the late-antique original. It also provided more complete versions of sermons that were already known and – until the moment of the discovery – considered intact. François Dolbeau, in a very important article, has calculated the effect of the transmission context on the length of the Augustinian sermons on Epiphany and Saints, concluding that a sermon of fewer than 833 words is almost certainly an abbreviation and that sermons of over 2,000 words are perfectly normal for Augustine.Footnote 15
Still, these markers should not be relied on too blindly. Even some sermons over 2,000 words in length turned out to be abbreviated when the Mainz codex offered up the more complete version. Similarly, the antiquity of a collection is no guarantee. Even in very old manuscripts, such as the main witness of De Paenitentia, abbreviated versions of sermons are found. The idea that the context of the transmission influences the likelihood of a sermon’s text being manipulated seems solid, but we must keep in mind the spans of time – in some cases centuries – for which we do not have extant manuscript witnesses and in which a lot of this manipulation may have happened.
The corpus of sermons wrongly associated with Augustine – and by extension the corpus of misattributed patristic preaching – is a largely untapped resource for the study of (early-)medieval religious history, if we can categorize and access it properly. An additional advantage of blurring the division between authentic and inauthentic Augustinian sermons is that it would give greater visibility to the creative medieval adaptations of Augustinian sermons that are currently mostly dismissed as degradations.
Editions, Reference Tools, and Heuristic Instruments for the Sermones ad populum
Given the heterogeneity of Augustine’s Sermones ad populum and the complexity of their medieval transmission, a few words about how to access this corpus can be of use.Footnote 16 The scholarly tradition initially focused mainly on the goal of distinguishing authentic Augustinian sermons from the vast corpus of misattributed or newly composed medieval sermons. This goal, combined with the practice of the early editors, has created a situation in which the Sermones ad populum were lifted out of their medieval collections and reintegrated into a new system of categorization, debuted by the Maurist editors, whose 1683–1684 edition is still widely consulted through its incorporation in the Patrologia Latina, volumes 38–39.Footnote 17 The Maurist system is constructed as follows:
1. sermones de scripturis (sermons on Scripture): 1–183;
2. sermones de tempore (sermons for specific times of the year and feast days): 184–272;
3. sermones de sanctis (sermons on saints): 273–340;
4. sermones de diversis (sermons on various subjects): 341–363;
5. sermones dubii (sermons of doubtful authenticity): 364–394.
To this was added a series of authentic homiletic fragments as well as an appendix containing inauthentic items. All subsequent discoveries or reinterpretations of sermons were introduced into the Maurist system and are referred to by means of a number followed by a letter. The largest collection of post-Maurist additions to the Augustinian homiletic corpus, 138 items, is printed in Morin Reference Morin1930. The Maurist numbering system has also directed the new critical edition of the Sermones ad populum in the Series Latina of the Corpus Christianorum, in which five volumes have been published at the time of writing.Footnote 18 The corpus in its entirety is accessible in an English translation by Edmund Hill (Hill Reference Hill and Rotelle1990–1997). It should be noted that Hill worked from the editions he had at his disposal, and the vast majority of them were not modern critical editions. Given the relative frequency of new discoveries and the fact that smaller contributions of a single sermon or even a fragment of a sermon to the edition project are regularly featured in journals, edited volumes and Festschriften, systematic overviews of the Sermones ad populum are indispensable. For Augustine’s Sermones ad populum Dolbeau Reference Dolbeau2019–2021 provides a recent update to Gryson Reference Gryson2007. For sermons erroneously attributed to Augustine, Machielsen Reference Machielsen1990, which covers pseudo-Augustinian sermons under item numbers 450–3387, is a precious resource. Digital repertories such as the Library of Latin Texts (LLT),Footnote 19 the Clavis Clavium (ClaCla),Footnote 20 and the PASSIM databaseFootnote 21 are increasingly replacing printed overviews. These databases have the advantage of greater flexibility compared to the printed reference works, in terms of search options as well as intermediate updates, but only if they are regularly updated and sustainably stored.
Conclusion
The medieval transmission of Augustine’s preaching, in particular that of the Sermones ad populum, has had a significant impact on which parts of his vast homiletic corpus have survived to our times and what state the texts find themselves in after over a millennium of being copied by medieval scribes. The history of the sermons’ transmission plays a part in the large diversity and disparity we observe in Augustine’s sermons, in terms of length, complexity of the argumentation, rhetorical features, and style. The sermons’ relative brevity, their lack of authoritative titles or organization, and above all their active reuse in medieval religious contexts means they were particularly prone to modification as part of their medieval reception and dissemination. An awareness of this fact when using the sermons as primary sources for scholarship is vital. Though it is common knowledge that the extant corpus of Augustinian sermons constitutes just a small part of his original production – some even say that no more than 10 percent of what he actually preached survives – the survival of the sermons that did make it to our times was neither wholly random nor wholly deliberate, and it was definitely not a matter of Augustine’s own intentions. There were accidents of transmission, sermons lost because of the destruction or disintegration of the manuscripts that carried them. There were many deliberate choices: to copy a certain sermon, to make it part of a much-used collection, to change its status from homily to treatise. On the whole, however, we cannot capture the fate of Augustine’s sermons in one set of rules. There is no teleological, clearly delineated system behind the corpus of sermons that survives, and certainly not one that originated with the author of the sermons himself. Hence any comments on the nature or characteristics of the sermon corpus as a whole, or of specific subsets within it, must be made with caution. However, this cautionary tale also has a positive twist. A more systematic and elaborate incorporation of information on the transmission of Augustine’s sermons in our research opens up a fascinating world of questions on the dynamic and intense medieval reception of the homiletic production of the patristic era.
