The PASSIM Project (Patristic Sermons in the Middle Ages): Towards a Virtual Research Environment for the Study of Patristic Sermon Collections

Riccardo Macchioro
Thousands of manuscripts witness to the huge circulation of Latin Patristic sermons throughout the Middle Ages, and it is hard to find two of them that are completely identical. [1] Homiliaries exhibit large variety with regard to size of the collection, codicological features, arrangement, sermons selected, author attribution, a piece’s textual shape, and so on. To orient themselves in such a vast literary production, scholars already have indispensable tools at hand – above all Eligius Dekkers’s Clavis Patrum Latinorum, Johann Machielsen’s Clavis Patristica Pseudoepigraphorum Medii Aevi, and Roger Gryson’s Répertoire Général, [2] all of which fulfilled the need for an extensive inventory of texts and authors. The enormous amount of material to manage prevented Dekkers, Machielsen, and Gryson from undertaking any thorough survey of the sermons’ manuscript transmission, requiring them to ground their work on the available printed editions. Practicality was prioritized at the expense of reliability and nuance. For the sermons in particular, this choice constitutes a major problem due to the quality of editions: in many instances, only the outdated editions published in the Patrologia Latina (PL) were (and still are) available. [3] The PL editions are philologically untrustworthy and often laden with false attributions. While they remain of invaluable importance, the Claves and the Répertoire portray patristic sermons as a fixed corpus cast in late antiquity, whereas they have always been manipulated, rearranged, and combined together, a process of reception that gave birth to countless different redactions and anonymous pieces. [4] It is hard to obtain from current repertories a full picture of the cross-referentiality and relationships that exist between all these sermons because information is scattered over various entries and often incomplete.
In order to tackle these problems, the PASSIM project and research tool adopts a novel change of perspective: it seeks to investigate patristic sermons as born in late antiquity but very much alive in the Middle Ages, and destined to widely influence medieval culture through their manuscript transmission. Two characteristics of PASSIM support its aims. First, the project’s pivotal focus is on reception, as witnessed by the medieval tradition. Second, PASSIM draws the information directly from the actual manuscripts. Of course, our current ability to make use of digital tools and data provides for a remarkable enhancement of the study of the medieval afterlife of patristic homilies: PASSIM supplies and makes searchable a great deal of data and metadata on patristic sermons and their manuscripts, all linked to each other in a knowledge network, thus fostering overarching investigations on this domain. The accompanying poster presents the features of the PASSIM VRE by means of a telling case-study: Augustine of Hippo’s sermon no. 247 (AU s. 247). [5]
A patristic sermon can be examined from different angles. One perspective is the original sermon as it was written (or preached) by a late-antique preacher, whose text and production circumstances have to be reconstructed philologically. Another is the available edition, a text that has been printed and that has become a reference point. It might aim at reconstructing the original – as in the case for modern critical editions – but is often not much more than the reproduction of one of the various shapes it takes in the manuscripts. The third view we might take is that of its medieval reception, an approach that makes every individual copy of a given sermon relevant. The particular features of the sermons in a collection reveal historical and cultural patterns and the interests and attitudes of preachers and users in specific times and places.
PASSIM bridges these three perspectives by featuring three different layers of representation for sermons. The “sermon details” section displays the piece exactly as it is found in a given manuscript, accounting for any deviation from what might be considered the standard (e.g. wrong attributions, variant readings in incipit and explicit, changes in feast-day assigned, etc.). [6]
Figure 1: Sermon details PS-FU s. 40
The scholarship available on each sermon is embodied through the “GoldSermon” (SG) function: the representation of a sermon as it is supplied by one of the reference editions, with the data (attribution, incipit, explicit, etc.) assigned to it there. At this level, PASSIM makes the most of the crucial reference system provided by the Claves and the Répértoire since each SG is labelled accordingly. Also, all the sermon manifestations are assigned the same labels, as long as they fit reasonably with them. [7] Finally, the third and most important level is provisionally called “SuperSermonGold” (SSG) [8] and represents an “authority record”: an ideal or standard version, as verified through scholarly means, of a sermon that may or may not coincide with one of the available editions. It collects all the identification codes that have been supplied by previous research (if applicable), and, as a consequence, all the sermon manifestations labelled with those codes will automatically attach to it. To each SSG, therefore, is at once linked an archive of all editions and scholarship on that sermon (making it possible to search also using attributions or reference numbers that were once in vogue) and its manuscript witnesses. This “authority record” is PASSIM’s foremost innovation in inventorying sermons, because it works as a hub for all the information about a sermon, placing it at the center of its philological, cultural, and historical network.
Heretofore, the field of patristic sermons has not been widely explored by digital projects. Valuable contributions have been provided mainly by endeavours aimed at the digitization of extant material, in terms of both manuscripts and texts of the sermons. [9] In addition, the system of references contained in the Claves is searchable through the Clavis Clavium Database ( Although such tools are undoubtably of help for research, none of them brings forth a real epistemological upgrade. They are still grounded in the edited and catalogued material, without encompassing the largely overlooked medieval manuscript transmission; they do not envisage the possibility of querying this material for the sake of a better grasp of the medieval reception of sermons; and there is no attempt to address and make searchable the intricate textual and transmission networks that underlie the whole corpus. An attempt to apply a manuscript-based approach to the medieval circulation of patristic sermons is being carried out by the Repertorio degli Omeliari del Medioevo (R.O.M.E.) project: R.O.M.E. indeed provides detailed descriptions of sermon collections (mostly, but not only, in Beneventan script), integrated in a vast and multifaceted digital archive for the study of medieval culture. [10] In this case, however, a limit is set by the relatively circumscribed number of manuscripts considered (and, as a consequence, of the amount of material available to the user for investigation).
PASSIM strives to offer a solution to all these issues: let us look at AU s. 247 SSG (the center screen in the poster) to understand how. The first fields provide the basic information on this standard sermon. Some regard the sermon itself: author (Augustine), incipit, explicit, keywords (an internal labelling system that enables the retrieval of items that share the same characteristics). The personal dataset, collections of items created by the user to be exploited for advanced queries, aids in building customized research records; the field indicating “historical collection(s)” (selections of homilies that circulated relatively consistently in several medieval copies, although each copy was always customized in a specific way), if applicable, situates the SSG in its transmission context.
Then the “Equality Set” follows, a key section for the goals of PASSIM. The field “Contains” collects all the SGs that “belong” to the SSG: this is the aforementioned “archive of editions.” Far more often than expected sermons have been edited twice (or more) with different attributions. AU s. 247 (see at the left-hand side of the AU s. 247 SSG page), for example, was published and ascribed to Augustine in PL 38, but appeared also in PL 65 under the name of Ps-Fulgentius (PS-FU s. 40). Each of them received its own set of labels in repertories, but only the attribution to Augustine is “true”: once recorded, selecting the “Elevate” button ensures that the one that proves to be the verified version will obtain the status of “authority record.” [11] The “Links” field connects this SSG with those sermons that share textual passages with it (“partially equals”), directing the user to the respective SSG pages of connected sermons. In this instance, we see here Caesarius of Arles’s sermon 175 (CAE s. 175) and Pseudo-Augustine’s sermon 169 (PS-AU s. 169), reproduced in the right-hand middle space in the poster. Both sermons employed AU s. 247 as a source. A further connection is revealed by looking at their respective Equality sets: they share another source, Augustine’s Epistula 137.
Finally, the “Manuscripts” section is another research innovation, representing PASSIM’s manuscript-based approach to the reception of patristic sermons. All the manuscripts linked to the items that feature in the “Equality Set” are connected to the SSG page, thus providing a complete picture of its transmission network. The list is made of links that direct users to a full description of each codex. Here every individual, concrete manifestation of the “ideal” standard (the SSG authority record) is located – and can be visualized – within its specific context, including the feast to which it is assigned, the arrangement of the homiliary, the series of sermons that surrounds it, its peculiar textual variations in incipit and explicit, and so on.

Figure 2: Ms. description (partial) – Wolfenbüttel, HAB, 4096.
In other words, it is possible to easily navigate the multifaceted reception of a sermon; having at hand all the individual manifestations of the sermon in each manuscript, the user is able to trace the particular traits of its reception in diverse milieux and times.
The manuscript St.-Mihiel Z.20 provides a sample for the poster (“Manuscript details,” in the bottom left corner of the poster’s central area). After a thorough codicological description, including links to external sources if applicable, every sermon is presented as it appears on the codex. Incipit, attribution, rubrics, paratexts, and feast days are accurately recorded as they appear in that specific manuscript (even if different from those found in the reference works or in the critical edition). This information is crucial to grasping the actual reception of a sermon in the particular milieu where that manuscript has been realized. St. Mihiel Z.20 was used to create the edition of PS-FU s. 40 (see item no. 42), but it is now clear that the name “Fulgentius” does not show up anywhere in the manuscript. The existence of Ps-Fulgentius is simply a scholarly hypothesis with no basis in the manuscripts. PS-FU s. 40 proves to be a ghost sermon that exists only in scholarly literature! It has always (and only) been Augustines’s sermon. This case demonstrates how, by turning back to the manuscripts and clarifying the distinction between manuscript evidence and scholarly tradition, the status quaestionis available through repertories can be significantly improved.
While in reference works all this data is scattered over various entries and is sometimes defective or misleading (with the manuscript tradition only sometimes and cursorily described), [12] the SSG centralizes the information on one interactive page: networks, reference codes, and manuscripts are displayed in a single view. More importantly, this will eventually be the case for thousands of sermons and manuscripts: each SSG page contributes to a comprehensive framework of relationships which enables virtually infinite navigation potential for the user. Every component is clickable and searchable, and every data typology is usable as a basis for queries. Thanks to the careful codicological and content descriptions, the perspective can shift from considering single sermons and their network on their own, to embracing entire manuscripts (the homiliaries) as cultural and historical objects, resulting from attentive planning but also determined by the shape of the sermons that were available to the compiler. As a next step, the user can turn to look at series of homiliaries (e.g. selected by provenance or belonging to the same monastic environment) or perhaps go back to the network of individual eye-catching sermons encountered while surveying the homiliary.
Figure 3: Mss. search window.
To be efficient, of course, PASSIM requires a huge amount of data, which is imported regularly directly from available digital libraries and from a wide range of libraries catalogues; it is entered manually in the case of especially significant manuscripts. Every secondhand piece of information undergoes an assessment so as to ensure its correctness and quality, and at all times a traffic light system makes the user aware of the status of the information being displayed (“Status” field: red = unchecked; yellow = provisional; green = verified).
Figure 4: Provisional content overview.
The section at the bottom of the poster shows a selection of the research domains that benefit from PASSIM. One is reception studies: the opportunity to handle and navigate a great deal of material discloses unexpected research avenues. In this case, Augustine’s Easter Sermon collections ended up being complemented by CAE s. 175, but how and why did this happen? Another research avenue is intellectual networks: shared material, recurrent series of homilies, and shared innovations foster exploration. PS-AU s. 169 and CAE s. 175 share sources; what can be discovered by investigating the reception of the Augustinian material involved? A final research domain that we engage is critical editions. For PS-AU s. 169 (poster: bottom right corner of the central screen), scholarship has always relied on the single-manuscript edition in the Patrologia Latina; however, it now turns out to be transmitted by several other manuscripts. The need for a new critical edition becomes immediately evident, and the PASSIM project immediately provides material to tackle the task. But PASSIM is not limited to these areas since the project strives for further enhancements in the field, envisioning developments in cooperation with other projects (some of which are already on their way), the implementation of more complex query possibilities on the database’s material, and improved strategies of visualization for the results of the inquiries conducted by means of the PASSIM research tool.
Figure 5: experimental “distance” visualization of relationships among clusters of connected sermons.
According to a rather strong scholarly attitude in the past decades, investigation into patristic texts falls into the field of Classical and Late Antique Studies, while their medieval reception falls into that of Medieval Studies. In more recent years, however, awareness that the two approaches largely benefit from each other has constantly increased: [13] one of the foremost contributions we expect from PASSIM is to further help in bridging these perspectives.


Boodts, S. 2019. “Navigating the vast tradition of St. Augustine’s sermons: Old instruments and new approaches.” Augustiniana 69(1):83–115.
Dekkers, E. 1995. Clavis Patrum Latinorum. Turnhout.
Dolbeau, F. 1990. “Sermons inédits de S. Augustin dans un manuscrit de Mayence (Stadtbibliothek, I 9).” Revue des Études Augustiniennes 36:355–359.
Étaix, R. 1994. Homéliaires patristiques latins: recueil d’études de manuscrits médiévaux. Paris.
Gryson, R. 2007. Répertoire général des auteurs ecclésiastiques latins de l’antiquité et du Haut Moyen Âge. Freiburg.
HÜWA. 1969–. Die handschriftliche Überlieferung der Werke des heiligen Augustinus. (Veröffentlichungen der Kommission zur Herausgabe des Corpus der Lateinischen Kirchenväter). Vienna.
Machielsen, J. 1990. Clavis Patristica Pseudoepigraphorum Medii Aevi. Vol. 1, Opera homiletica. Turnhout.


[ back ] 1. “Patristic Sermons in the Middle Ages. The Dissemination, Manipulation, and Interpretation of Late-Antique Sermons in the Medieval Latin West,” funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program under grant agreement No ERC-2018-stg 802210; Radboud University Nijmegen (Department of History, Art History and Classics – RICH, Radboud Institute for Culture and History).
[ back ] 2. Gryson 2007; Dekkers 1995; Machielsen 1990.
[ back ] 3. Often reproducing even earlier print publications, such as the Maurists’ edition of Augustine’s Opera omnia in the seventeenth century.
For a much more detailed account of these scientific challenges, and on the significance of the medieval afterlife to fully grasp the nature of sermons, see Boodts 2019.
[ back ] 5. I choose to focus here on the heuristic value of PASSIM, rather than on its functionalities or technical features, which are very briefly summarized as follows: the PASSIM web application is built on the Django framework, the open-source web framework based on Python designed to simplify the creation of complex database-driven websites. The data are stored in an SQLite database.
Seminal works have demonstrated the importance of reception (Étaix 1994) and showcased that an extensive examination of medieval manuscripts with patristic content can still lead to astonishing findings. To mention just one, see Dolbeau 1990.
[ back ] 7. If not, in fact, it means that this is a sermon that has not been identified previously, or at least that it differs so much from the edited piece(s) that it has to be regarded as another text.
[ back ] 8. Names such as “GoldSermon” and “SuperSermonGold” are just unofficial, provisional labels created by PASSIM to identify these new kinds of classification. The sense of the terms is that they correspond to increasing layers of reference standards. “Gold” is a standard record, firmly established (and widely employed), though still enchained to printed edition and often not reliable. “Super” is a new standard that upgrades the conception itself of the reference: a real “authority record,” meant not only to be definitive but also to embody an entire archive of digital information.
[ back ] 9. The sermons published in the PL are accessible through the Patrologia Latina Database ( and through Corpus Corporum (, while the publisher Brepols makes available through the Library of Latin Texts ( the modern (critical) editions published in its Corpus Christianorum series.
[ back ] 10. This is the Digital Archive for Medieval Culture “Mirabile” (, where R.O.M.E. itself is hosted) of the International Society for the Study of Medieval Latin Culture (S.I.S.M.E.L., Florence).
[ back ] 11. This is not always feasible: there are still plenty of sermons with multiple attributions, for which it has not yet been possible to determine the true one. The important point is that the entire network of connections (which is a matter of fact) remains in place in any case.
[ back ] 12. Machielsen 1990 sometimes provides some information on manuscripts; most relevant in this respect is also the HÜWA project (Die handschriftliche Überlieferung der Werke des heiligen Augustinus), which through a series of printed volumes (still in in progress) aims to inventory all the manuscripts that contain works by Augustine.
[ back ] 13. See Boodts 2019:85–91.