Augustine vs Wodan

Introduction

In Roman Late Antiquity, the seven-day week outrivalled its main competitor of eight days ^[1]. The rise of Christianity in the fourth century, from persecution to state religion, was instrumental in this process. One of the pillars of the new religion was the Judeo-Christian creation story in Genesis with its six days of work and one day of rest. In the scripture, these were simply numbered, and this led to the standard medieval denomination system of feria 1 ( or dies dominicus/a ) = Sunday to feria 7 ( or sabbatum ) = Saturday, which is still represented in Portuguese, Mirandese, and Gallician:

The Romans, following Hellenistic tradition, preferred to name the seven days after the seven planets, which in turn received their labels from Roman gods: Mars, Mercury, Jupiter, Venus, Saturn, and sun and moon. These remained popular throughout the Middle Ages and can still be found in most Romance languages like French, Spanish, Italian, or Romanian:

When Christianity, and with it the written word and Latinity, spread to the Germanic peoples across the Rhine and in Britain, Roman gods soon gave way to local ones: Tyr, Wodan, Thor, Freya ^[2]. These are still represented in modern Germanic languages, like English, German, Dutch, or Danish:

One weekday stands out in this table ( besides Danish lørdag, the Scandinavian ‘washing-day’ ), German Mittwoch, ‘middle of the week’. Scholarship tends to explain the reason for this shift from Wodan’s day to midweek by focusing on the Germanic pantheon ^[3]: Wodan sat at its top, and his cult was and remained so influential that ecclesiastical elites around the turn from the first to the second millennium AD felt the urge to act and wipe his name from daily references; the Scandinavian mission and rivalries between certain cults – Wodan vs Tyr – may have fed into this process. What these interpretations do not appreciate enough is the institutional setting of the first written occurrence of this term, mittauuechun, by Notker Labeo of St Gall in his psalms commentary of around the year 1000. This essay will explain Notker’s motivation for replacing Wodan’s day through St Gall’s Irish scholarly tradition.

St Gall has often been considered a prime example of Irish intellectual influence on the Continent, not least because of the survival of some fifteen Irish script codices ( or parts thereof ) among the St Gall holdings, a unique ninth-century list of books written in Irish script, numerous Irishmen among the St Gall name-lists, and the famous story of the Irish bishop Marcus and his nephew Móengal / Marcellus remaining at St Gall with their book and some retinue, with Móengal / Marcellus becoming the schoolmaster who taught the golden generation of St Gall scholars ^[4]. Two scholars in particular, the St Gall librarian Johannes Duft in the 1950s and more recently Sven Meeder in 2018, have questioned this interpretation in some detail ^[5]. They stress, quite rightly, that one needs to be careful not to overstate the physical presence of Irish monks at St Gall and therefore should avoid the term ‘Irish colony’. More problematic, however, is their view that Irish script books brought to St Gall had no discernible impact; these were considered ornamental at best, but of no use to the community. Meeder additionally argues that there is no evidence that St Gall received Irish texts directly from Ireland. The following analysis will demonstrate, en passant, that both assumptions cannot stand.

By focusing on books and texts, these studies have not seriously considered the potential influence of Irish ideas on St Gall intellectual culture. St Gall had two Irish schoolmasters, the above-mentioned Móengal / Marcellus in the third quarter of the ninth century and the less well-known Fáelán towards the end of the tenth ^[6]. One would expect that their teaching left some mark on the community. This article argues that Notker’s replacement of Wodan’s day with the more neutral ‘middle of the week’ owes to these Irish intellectuals. Identifying the transmission and impact of ideas, especially if conveyed orally, is not a straight-forward task and by definition has to be based on circumstantial evidence. The chain of evidence provided here is as follows: First, replacing Wednesday’s reference to a pagan deity with a more neutral term, and in this following Augustine’s commentary on Psalm 93, happened before Notker Labeo only in Ireland, in a computistical text of 718/719. Second, Móengal / Marcellus had a strong interest in both areas underlying this replacement, the study of the psalms and their commentaries on the one hand and computistical literature on the other; the charter evidence even demonstrates that Augustine’s commentary on Psalm 93 underlay his thinking and that week-day terminology was part of his teaching. Third, Fáelán, a century later, continued Móengal / Marcellus’ teaching, particularly of computistica, from a specific Irish angle; his floruit coincides with the young Notker Labeo, who shared Fáelán’s interest in computistica and was, in this field, particularly fascinated by Irish ideas; he introduced ‘middle of the week’ for Wednesday in his Old High German discussion of Psalm 93, which almost exclusively used the Augustine commentary acquired under Móengal / Marcellus.

The Munich Computus of 718/719 and the Reception of Augustine’s Commentary on Psalm 93

Notker and German speakers were not the first to substitute the attribution to a pagan God of the fourth day of the week with a more neutral term. The earliest example comes from an Irish textbook on the reckoning of time. Seventh-century Ireland designed a new Christian curriculum of principally grammar, exegesis, and Christian time-reckoning ( computus ), but also history and music ^[7]. Computus fulfilled primarily the basic need of structuring the liturgical calendar that revolved around the moveable feast of Easter, but it also had more exegetical overtones in explaining God’s creation as manifested in the cosmos. The insular world of the early eighth century produced four extant comprehensive textbooks on this calendrical science, of which Bede’s ‘De temporum ratione’ of 725 soon eclipsed the earlier Irish ‘Computus Einsidlensis’ of ca. 700 x 718, the ‘Munich Computus’ of 718/719, and ‘De ratione conputandi’ of the 720s ^[8].

The ‘Munich Computus’ is the only one of these Irish texts that can be dated with certainty, as it incorporates a dating clause to the present year of the author ^[9]. It applies the standard framework of such textbooks, discussing solar theory first, followed by lunar theory and finally the combination of the two, Easter calculations. Solar theory was structured according to the 14 divisions of time, from the smallest, the indivisible atom, to the largest, the more exegetical mundus ( ‘world’ ) that comprised all time. In this last category, the principal focus of the Munich Computist is to demonstrate the chronological parallels between three core events that define the Christian religion: Creation, Exodus, and Crucifixion / Resurrection. The Munich Computist here stands in a tradition of commentators on the question of dating Easter, as similar discussions can be found in the fourth-century ‘Cologne Prologue’ to the 84-year Easter cycle called the Supputatio Romana ^[10], and Victorius of Aquitaine’s Prologue to his 532-year Easter cycle of 457 ^[11]. The Irish authors worked directly from Victorius’ account, but modified it to achieve identical chronological sequences for all three events ^[12]. For this, it was essential to define a day as extending from midday to midday, a custom the author attributes to the Hebrews ^[13]. The Munich Computist’s schema may best be summarised in table format:

Of interest for the present study is the weekday terminology used throughout this comparison of Biblical weeks. In an earlier chapter, the Munich Computist provides a brief summary of the development of weekday terminology, from the Hebrews through Roman planetary weekdays to the consecutively numbered feria system whose introduction is here attributed to Pope Silvester ^[14]. When discussing the three Biblical weeks, however, the author employs what appears on the surface to be some sort of a hybrid system: The Christian dies dominicus ( ‘Lord’s day’ ) is followed by Roman planetary dies lunis ( ‘day of the moon’ ) and dies Martis ( ‘day of Mars’ ), then the bilingual Latin / Old Irish dies cetene ( ‘day of fast’ ), and finally the Christian feria V, feria VI, and sabbatum. The Old Irish names of the week, largely attested in eighth- and ninth-century glosses, are a mix of planetary and Christian terms: dia domnaig ( ‘Lord’s day’ ), dia luain ( ‘day of the moon’ ), dia máirt ( ‘day of Mars’ ), dia cétáine ( ‘day of first fast’ ), dia dardáin ( ‘day between the fasts’ ), dia oine didine ( ‘day of last fast’ ), dia sathairnn ( ‘day of Saturn’ ) ^[15]. Modern commentators on the Munich passage therefore explained bilingual dies cetene as a slip by the Irish author of this text, unconsciously reverting back to his native tongue ^[16].

There are two caveats in this theory. First, this is not an isolated example of Old Irish in the ‘Munich Computus’, and in the other two instances the purpose appears to be decidedly didactic ^[17]. This is further corroborated by the use of Old Irish in the closely related, slightly earlier ‘Computus Einsidlensis’ of the early eighth century, where its didactic function rooted in class-room teaching is even more apparent ^[18]. Second, if dies cetene was a subconscious slip, one would expect this to have occurred only once; but the entire sequence of planetary weekdays for Monday and Tuesday, dies cetene for Wednesday, and Christian terminology for Thursday to Sunday is repeated three times in this passage. This is a deliberate and systematic design, and dies cetene fulfils a very specific didactic function in it. The decisive clue is presented by the author himself: When introducing this passage, the Munich Computist does not reference Victorius of Aquitaine – whose Prologue underlies this discussion – as the main source for this passage, but Augustine. The bishop of Hippo was considered core reading for any early medieval exegete interested in Biblical events. Besides his more famous texts, especially his Genesis commentaries, one could find additional information in less obvious locations, like his commentary on the psalms. Psalm 93 deals with the fourth day of the Hebrew week ( quarta sabbati ). In his commentary, Augustine provides some details on the first four days of Creation that would have been immediately relevant to the Munich Computist trying to make a case for the chronological parallelism of the weeks of Creation, Exodus, and Crucifixion / Resurrection. Concerning the title of this psalm, Augustine has the following to say ^[19]:

Quare ergo talem habet titulum: in quarta sabbati? Una sabbati, dies dominicus est; secunda sabbati, secunda feria, quem saeculares diem Lunae uocant; tertia sabbati, tertia feria, quem diem illi Martis uocant. Quarta ergo sabbatorum, quarta feria, qui Mercurii dies dicitur a paganis, et a multis christianis; sed nollemus; atque utinam corrigant, et non dicant sic. Habent enim linguam suam qua utantur. Non enim et in omnibus gentibus ista dicuntur; multae gentes aliae atque aliae aliter atque aliter uocant. Melius ergo de ore christiano ritus loquendi ecclesiasticus procedit.

Why, then, does it have such a title: “on the fourth of the Sabbath”? “One of the Sabbath” is Sunday; “the second of the Sabbath” is Monday, which the pagans call “the day of the moon”; “the third of the Sabbath” is Tuesday, which those [ i. e. the pagans ] call “the day of Mars”. “The fourth of the Sabbath”, then, is Wednesday, which is called “the day of Mercury” by the pagans, and by many Christians; but we do not want this: may they [ the Christians ] correct it, and let them not call it thus. In fact, they have their own language, which they should use. In fact, these [ days ] are not called thus among all people; many different people call them in different ways. An ecclesiastical manner of speaking certainly comes better out of a Christian mouth.

On the surface, this is principally an emphatic statement against the use of Roman planetary weekdays to avoid association with pagan gods. More literally, however, it specifically urges the reader to change the terminology of the fourth day of the week, the subject of this psalm, to a more neutral or Christian vernacular reading. This is exactly what the Munich Computist does by substituting dies Mercurii with the bilingual dies cetene ( ‘fast-day’ ). In this context, the hybridity of terminology in the Munich passage can be read as a didactic tool for conveying Augustine’s message: The same weekdays as listed by Augustine are used first ( dies dominicus, dies Lunae, dies Martis ); then, Augustine’s advice is followed by substituting dies Mercurii with a more neutral vernacular term, before continuing with Christian feria terminology. For a student having discussed Augustine’s psalm commentary in exegetical class, the didactic message cleverly woven into this passage is clear: there are Roman planetary and Christian feria weekdays; remember Augustine’s psalm commentary on the fourth day of the week, in which he called for the use of a vernacular term for this day; Christian terminology is to be preferred over the Roman planetary ones referencing pagan gods. More generally, this Augustine reception provides an explanatory framework for the origin of the Old Irish hybrid system of planetary and Christian weekdays.

Móengal / Marcellus’ Intellectual Influence at St Gall

The ‘Munich Computus’ had a notable reception in the two centuries after its initial composition ^[20]. Its only surviving copy was produced at St Emmeram in Regensburg in the late 810s or early 820s ^[21]. The earlier ‘Computus Einsidlensis’ from which the Munich Computist worked, however, left hardly any traces on the Continent ^[22]. Its only surviving copy was written at St Gall at some point between 874 and 892. The survival of numerous insular abbreviations in this Continental copy of the text suggests that it was produced directly from an exemplar in an insular hand, which makes it likely that the ‘Computus Einsidlensis’ reached St Gall directly from Ireland shortly before 874 x 892 ^[23]. This links the arrival of this text to a famous episode related in the ‘Casus sancti Galli’, the house chronicle of the St Gall monastic community.

These ‘Fates ( of the monastery ) of St Gall’ were originally composed by the monk Ratpert, who covered the period from St Gall’s foundation in 614 to his time of writing in 884 ^[24]. This was continued to 972 by Ekkehart IV, at some stage after his return to St Gall from Mainz in 1034; the fact that the narrative breaks off more than half a century before the time of composition makes it likely the author’s death in the late 1050s was responsible for the unfinished state of this work. Ekkehart covered six abbacies in full ( Salomo III, Hartmut, Engilbert, Thieto, Craloh, Purchat ) and ends in the second year of Abbot Notker ^[25]. His approach differed from Ratpert’s in that he moved the focus considerably from politics to culture, from the St Gall abbots to other outstanding members of the community, especially its librarians, schoolmasters, scholars, and artists. This becomes evident immediately at the beginning of his work on the abbacy of Salomo III. The narrative starts by placing Salomo among a group of outstanding scholars and artists of his generation, active in the last quarter of the ninth century: Notker Balbulus, a musical scholar, gifted poet, and biographer of Charlemagne; Tuotilo, a famous musician and sculptor, especially of ivory; and Ratpert, the first author of the ‘Fates of the monastery of St Gall’ and head of its monastic school ^[26]. They were able to fulfil their potential only through the teaching, guidance, and influence of the schoolmaster Iso and the Irishman Móengal. It is worth citing Ekkehart’s account of Móengal in full ^[27]:

Grimaldi temporibus canonici abbatis, Hartmuoto eius quasi proabbate, Marcus quidam Scotigena episcopus Gallum tamquam compatriotam suum Roma rediens visitat. Comitatur eum sororis filius Moengal, postea a nostris Marcellus diminutive a Marco avunculo sic nominatus. Hic erat in divinis et humanis eruditissimus. Rogatur episcopus loco nostro aliquandiu stare allecto nepote. Diu secum deliberantes vix tandem consenserant. Dieque condicto partitur Marcellus nummos avunculi sui multos per fenestram, timens, ne discerperetur ab eis. Fremebant enim in illum, quasi ipsius suasu episcopus restaret. Equos autem et mulos, quibus ipse voluit, nominatim episcopus tradidit. Libros vero, aurum et pallia sibi et sancto Gallo retinuit. Stola tandem indutus abeuntes benedixit. Multis autem lacrimis utrimque discessum est.

Remanserat episcopus cum nepote et paucis suę linguę apparitoribus. Traduntur post tempus Marcello scolę claustra cum Notkero, postea cognomine Balbulo, et cęteris monachici habitus pueris; exteriores autem, id est canonicę, Ysoni cum Salomone et eius comparibus. Iocundum est memorari, quantum cella santi Galli his auspitiis crescere ceperit tandemque floruerit Hartmuoto eam, Crimaldi quidem vicario tandemque abbate, omnimodis augmentante.

At the times of Abbot Grimald, a secular canon, while Hartmut was quasi vice-abbot, Marcus, a certain Irish-born bishop, returning from Rome, visited Gallus as his compatriot. His sister’s son Móengal, later called by us Marcellus from his uncle Marcus by diminutive, accompanied him. This one was highly erudite in the sacred and the profane. The bishop was asked to stay in our place for a while, the nephew having already been won. Deliberating long among themselves, with difficulty they finally consented. And on the agreed day, Marcellus, being fearful, distributed many coins of his uncle through a window, so that he may not be torn to pieces by them [ their companions poised to leave ]. For they grumbled about him, as if the bishop stayed on his advice. Horses and mules the bishop personally gave to whoever he deemed worthy. Books, however, gold, and garments he kept for himself and Holy Gallus. Wearing the stola, he eventually blessed those leaving. Both parties, however, parted under many tears.

The bishop stayed with his nephew and some servants of their language. After some time, the cloister schools were given to Marcellus, with Notker, later with the epithet the Stammerer, and others of the monastic habit as students; the exterior, that is the secular canonical schools, however, with Salomo and his companions, were given to Iso. It is uplifting to be reminded how much the cell of St Gall began to grow under these auspices and eventually flourished under Hartmut, the deputy of Grimald and eventually abbot, while enriching the monastery in every way.

Ekkehart treats the arrival of Marcus and Móengal / Marcellus as a new departure in the cultural history of the monastery. Grimald’s abbacy started in 841, and Hartmut’s caretakership is usually linked to Grimald becoming arch-chaplain at the death of Baturich of Regensburg in 848 ^[28]. The charter evidence, however, suggests that Marcus and Marcellus had arrived at St Gall by 847 ^[29], and Hartmut may well have been caretaker for the busy Grimald at any point in the 840s. The St Gall necrology commemorated Marcellus, using both his Irish and his Latin name, as “most learned and most distinguished man” ( vir doctissimus et optimus ) under his death date of 30 September, no doubt a testimony to the difference this teacher made in the educational programme of this monastery ^[30]. Ekkehart’s description of Marcellus’ learning credits him broadly with expertise “in the sacred and the profane” – a reference to Christian and pagan literature –, which is then more clearly specified as comprising the artes liberales, and of these particularly music ^[31]. This is an echo of Notker Balbulus’ ‘Gesta Karoli’, which starts with an account of learning and patronage under Charlemagne ^[32]. Famously, the scene is set by two Irishmen – a certain Clemens and an unnamed monk –, who, “incomparably erudite both in secular and sacred writings” ( et in sęcularibus et in sacris scripturis incomparabiliter eruditos ) arrived in Francia publicly declaring that they would sell their wisdom for lodging and board. Slightly later in the text, Abbot Grimald of St Gall is anachronistically described as a pupil of Alcuin, who instructed him in the disciplines liberales. This story culminates in a special appreciation of music, the “modulations of chants” ( cantilenę modulationes ) ^[33]. Notker’s story confirms that Ekkehart’s eleventh-century description reflects the contemporary ninth-century perception. It will have helped that the developments and traditions initiated by Marcellus in the mid- to late ninth century were still strong in the eleventh. Certainly, in music, the discipline singled out in the two texts, the Lake Constance region was as much a pioneer and powerhouse in Notker’s time as in Ekkehart’s ^[34].

The more general mention of the artes liberales, however, may be more a statement of ambition or a reference to an idealised curriculum than necessarily a marker of factual school teaching and scholarly activity ^[35]. And it was not the only system of knowledge discussed among the early medieval literati. For our context, a division of disciplines referenced in Ermenrich of Ellwangen’s Letter to Abbot Grimald of St Gall is of particular interest. Ermenrich received his education at Fulda and is then attested at the court of Louis the Pious in the 830s, in the monasteries of Reichenau and St Gall in the 840s and 850s; from 866 until his death in 874, he was bishop of Passau ^[36]. In the early 850s, he composed a long treatise addressed to Grimald that may best be described as an open letter of ( self- )recommendation. His display of knowledge includes a division of philosophy into the three strands ethics, logic, and physics; of these, physics included the quadruvial disciplines of arithmetic, astronomy, geometry, and music, but also astrology, mechanics, and medicine ^[37]. This system of knowledge has a long tradition. Ermenrich likely got acquainted with it in his Fulda days, as Hrabanus Maurus was familiar with it through a set of schemata found in manuscripts containing Alcuin’s ‘De rhetorica’ and ‘De dialectica’ that also served as Ermenrich’s source ^[38]. Ultimately, however, as demonstrated by Bernhard Bischoff ^[39], this division of disciplines goes back to earlier Irish sources, as it can be found in the ‘Anonymus ad Cuimnanum’ ^[40], and also in ( pseudo? )-Isidore’s ‘Liber numerorum qui in sanctis scripturis occurrunt’ ^[41] and in Aldhelm’s writings ( influenced by his Irish teaching? ) ^[42].

This revived seven-fold division of physics may therefore have resonated well with Marcus and Marcellus, who arrived in St Gall a few years before Ermenrich wrote his letter; in fact, they may well be identified with the Irish scholars whose commission of a new Vita of St Gall Ermenrich hoped to claim for himself, and referencing this division of disciplines would then serve the purpose to demonstrate that Ermenrich was equally up to the task ^[43]. But, like the artes, this was, first and foremost, a theoretical construct often invoked in narrative sources. More indicative of what was considered relevant and important knowledge are, no doubt, the library holdings. For St Gall we are in the unique position that not only are many of its ninth-century books still in situ, but also that six ninth-century library catalogues survived. Their editor, Paul Lehmann, published these in what he considered their chronological order, written over the years ca. 850 to 896 ^[44]. Famously and uniquely, these start with “books written in Irish ( handwriting )” ( Libri Scottice scripti ) ^[45]. Lehmann may well be right in dating this list to the mid-ninth century, but there are also good reasons to consider it a later addition, probably at some stage in the 880s ^[46]. There is no compelling argument, however, not to assume that at least some of these books were part of Marcus and Marcellus’ travel library donated to St Gall.

The list of Irish books is quite substantial, comprising some 31 items. If this could be considered roughly the dimension of the book collection brought by these two Irishmen to St Gall ( and even if not all of these books were brought by Marcus and Marcellus, they evidently also brought others, as discussed below, so this number may well be representative ), it will have provided a substantial boost to the library holdings, which can be estimated – if surviving manuscripts are representative – at ca. 40 books around 816, and ca. 140 books around 840 ^[47]. The arrival of Marcus’ and Marcellus’ substantial travel library may therefore have been one of the triggers for a first inventory of the St Gall holdings as represented in the oldest library catalogue, the ‘Breviarium librorum’, which lists some 390 items; its core seems to have been written in the 850s, and was then constantly added to until the 870s ^[48]. Its structure is intriguing. It starts with Scripture, first Old and then New Testament books. This is followed by the church fathers ( here Gregory the Great, Jerome, Augustine, Ambrose ), authoritative ecclesiastical authors ( Prosper, Bede, Isidore, Cassiodorus, Eusebius, Libri diuersiorum auctorum, Alcuin ), and finally specific subjects ( monastic rules, hagiography, laws, glossaries, homilies, orthography, astrology, metrics, hymns ). To some degree, this hierarchical structure of knowledge by degree of ecclesiastical authority is very intuitive in a monastic setting ^[49], and St Gall will have taken some inspiration from the earlier Reichenau catalogue of 821/822 ^[50]. But, despite some loss of structure towards the end due to continuous additions, the St Gall catalogue arrangement is neater ^[51], and resembles more clearly the Irish system for seeking Christian truth as applied in Cummian’s Letter ‘De controversia paschali’ of the 630s: study the Bible first, then the church fathers, and finally more recent specialist literature ^[52].

If Marcus’ and Marcellus’ influence in the restructuring and systematisation of the St Gall library can better be sensed than grasped, their input in the ensuing acquisition process may be more apparent. Besides the main inventory, the ‘Breviarum librorum’, the remaining four catalogues record books acquired in the abbacy of Grimald with the help of Hartmut ( 841–872 ), books commissioned by Hartmut during his abbacy ( 872–883 ), and Hartmut’s and Grimald’s private collections. Of particular interest for the present study is the fact that the largest acquisition made by either abbot was a six ( ! )-volume copy of Augustine’s commentary on the psalms, and a two-volume excerpted short version ^[53]. Five of the six volumes of the full text still survive in St Gall under the modern shelf-marks Cod. Sang. 162–166, the sixth is now, incomplete, in the Zentralbibliothek of Zurich ( Car. C 32 ) ^[54]. Bischoff dates their script to ca. 850 ^[55], and the fact that they also appear in the main hand of the initial inventory, the ‘Breviarum librorum’ ^[56], confirms that they should be considered some of the earliest acquisitions made, shortly after the arrival of Marcus and Marcellus. The same holds true for a three-volume edition of Cassiodorus’ commentary on the psalms, which appears in Grimald’s acquisition catalogue and the ‘Breviarum librorum’ ^[57], and survives as Cod. Sang. 200–202, dated by Bischoff to the mid- or third quarter of the ninth century ^[58]. Natalia Daniel drew attention to the fact that both the six-volume Augustine and the three-volume Cassiodorus psalm commentaries show Irish palaeographical features ^[59]. That Marcus had a particular interest in the psalms is corroborated by the note in Grimald’s list that the abbot “gave a good psalter to the Irish Marcus, which was deposited in the church” ( Psalterium bonum Marco Hibernensi dedit, quod est positum in ecclesia ) ^[60], the only direct reference to Marcus or Marcellus in the St Gall catalogues. Though this appears not to have been Cod. Sang. 27 produced at this time ^[61] – a beautifully scholastic psalter, with the main text per cola et commata ( like Irish psalters ^[62] ) in the centre, and commentary in the extensive margins on the left and right –, the arrival of the Irish scholars at St Gall coincides with a new stage of psalter writing on the Continent that was initiated in this monastery in the mid-ninth century ^[63]. Marcellus can be identified as one of the named copyists in the mid-ninth century Greek psalter with Latin interlinear glossing ( Basel, Universitätsbibliothek, A VII 3, fol. 23r ) ^[64]. This codex belongs to a group of three Greek manuscripts with interlinear Latin translation, the other two covering the Gospels ( St Gall, Stiftsbibliothek, Cod. Sang. 48 ) and the Pauline Epistles ( Dresden, Sächsische Landesbibliothek – Staats- und Universitätsbibliothek, A.145.b ). A text by a Marcus monachus, at the end of the Dresden codex, with cross-reference under that name on fol. 36r, was considered by Traube and others to be the St Gall Marcus, but more recent scholarship has identified this as a fragment of ‘On the spiritual law’ by the early fifth-century Greek author Mark the Hermit ^[65]. Marcellus is also one of two St Gall scribes of a codex with New Testament texts other than the Gospels ( Zurich, Zentralbibliothek, C 57, fol. 73r ) ^[66].

This demonstrates a deep-rooted interest by Marcus and Marcellus in Biblical studies and especially the psalter. It may therefore not be a coincidence that their arrival at St Gall coincided with an unprecedented level of engagement with the psalter in that monastery as evidenced by the acquisition for the St Gall library of the main psalter commentaries ( including Augustine’s ), the design of a new scholastic layout for the psalter proper, and a comparison between the Greek and the Latin text. An insight into how an Irish schoolmaster like Marcellus would have taught the psalter, however, can better be gleaned from the heavily glossed copies of essential school texts. Irish scholars produced a number of these, most famously for the grammarian Priscian and for Bede’s scientific works. The most heavily glossed to survive, however, is the famous Ambrosian psalter commentary ( Milan, Biblioteca Ambrosiana, C 301 inf. ). It was written in the early ninth century, in Ireland or Bobbio ^[67], and may be representative of what Irish scholars considered important to convey in the classroom in psalter studies ^[68]. For Psalm 93, the explanation to quarta sabati in the Latin title starts with the Irish term for the fourth day of the week: dia cétaíne rogabad insalmo ł. Is dingnim forchomnaccuir is ind laithiu sin is immaircide .i. duchésad chríst ( “this psalm was sung on a Wednesday; or it is appropriate to the deed which happened on that day, to wit, to the Passion of Christ” ) ^[69]. This will have reminded the reader of Augustine’s advice to use the vernacular for this weekday to avoid reference to a pagan god.

It is quite remarkable that the main St Gall catalogues, the ‘Breviarum librorum’ and the acquisition lists, do not place much emphasis on schoolbooks, particularly in the ‘quadruvial’ or ‘physical’ disciplines. The personal inventories for Grimald and Hartmut, on the other hand, are full of them ^[70]. This is a good reminder that in any given monastery, books obviously also existed outside of the main library, in private collections, but presumably also on the shelves of the monastic classroom. If one is interested in a particular discipline, the wider monastic holdings, as long as they can be reconstructed, will provide a better and fuller insight than library catalogues. For the present study, the calendrical science of computus is of special interest ^[71]. St Gall certainly had a decent computistical collection at the time when Marcus and Marcellus arrived. The oldest manuscript of computistical content is the famous Cod. Sang. 225, written in 760 x 778, possibly 773 ^[72]. It combines in one volume grammar, computus, medicine, exegesis, and other subjects, and is very much reminiscent of schoolbooks before the Carolingian Renaissance, before individual disciplines were separated into dedicated books. Similarly, a scholar writing in insular script included computistica, some of which Irish, in a notebook of the late eighth century that included the famous ‘Vocabularius St Galli’ and that was soon incorporated in the St Gall holdings ( Cod. Sang. 913 ) ^[73].

A new departure was heralded by the momentous computistical year 816, for which the St Gall community imported latest calendrical theory from St Denis, no doubt facilitated by their former Abbot Waldo ( abbot of St Gall 782, of Reichenau 786, of St Denis 806 ) ^[74]. Two copies of this St Denis collection were made, which in subsequent years – before the middle of the century – were bound with what were to become the standard handbooks of the ninth century, Bede’s ‘De temporum ratione’ and Hrabanus Maurus’ ‘De computo’, respectively. The manuscript transmitting Hrabanus’ text, Cod. Sang. 902, falls into three parts, the ‘Aratus Latinus’, Hrabanus’ ‘De computo’, and the computistical collection from St Denis, which neatly fits the description in the initial St Gall inventory, the ‘Breviarum librorum’, of Liber astrologiae et computus Rabani et alius compotus in volumine I ^[75]. The Bedan manuscript, Cod. Sang. 251, is more interesting for our context. This copy of ‘De temporum ratione’ sporadically includes insular abbreviations, also in its glosses ^[76], and is closely related to a Bobbio ( ? ) copy of 836 x 854 ( Milan, Biblioteca Ambrosiana, D 30 inf. ) ^[77]. This may not be enough to argue that the St Gall text was copied from an exemplar in an Irish hand in Marcus’ and Marcellus’ possession, or even that either of them is one of the correcting and glossing hands ^[78]. But it certainly wetted the appetite of the St Gall librarian to acquire a full – and clean ^[79] – set of Bede’s scientific works ( ‘De natura rerum’, ‘De temporibus’, and ‘De temporum ratione’ ). Significantly, this was done through Irish channels at roughly the time of Marcus’ and Marcellus’ arrival, by the St Gall sister codex ( Cod. Sang. 248 ) in Caroline minuscule of the famous Karlsruhe Bede ( Karlsruhe, Badische Landesbibliothek, Aug. perg. 167 ) in Irish script with Old Irish glosses that Irish scholars brought from north-eastern France ( Soissons? ) to Lake Constance in the 840s ^[80].

This acquisition of essential computistical literature was surely supported by Abbot Grimald, who had a keen interest in computistica as evidenced by his vademecum ( Cod. Sang. 397 ^[81]; much of this material was later, towards the end of ninth century, integrated in a similar type of compilation, but larger in format, in Cod. Sang. 184 ^[82] ). A systematic review of the computistical holdings at St Gall, however, was only started after Grimald’s death by his successor Hartmut, whose private library not only contained core literature on the artes liberales, but also “Bede’s two booklets on the nature of things and on times; also a large one and other argumenta for calculation in one volume” ( Bedae libelli duo de natura rerum et temporibus; item unus grandis et alia argumenta computandi in volumine I ) ^[83]. The description of the second of these fits quite neatly the content of the St Gall codex ^[84] Einsiedeln, Stiftsbibliothek, 321 ( 647 ), pp. 27–156, which does not contain a major text of a known author. The Easter table covering 874–1101 suggests that the codex was written in its first 19 years, 874 x 892 ^[85]. This compilation draws on the existing St Gall holdings by incorporating material from Cod. Sang. 225 and 251/902, and it integrates the more practical chapters of Bede’s ‘De temporum ratione’ ( chs. 1, 3, 19, 23, 56 ). Literally the heart of this codex, however, is the Irish ‘Computus Einsidlensis’ on pp. 83–125. Significantly, the beginning of this text contains insular abbreviation ( particularly for autem, con, est, haec, per, vel ), but these peter out after the first few pages. Clearly, the copyist realised soon after starting his work on this text that the insular abbreviations may pose problems to a Continental reader. The conversion of these abbreviations was not always successful ^[86], but on balance the scribe was very careful, especially with the few Old Irish phrases and words scattered throughout the Latin text ^[87]. This evidence suggests that Hartmut was impressed by this Irish computistical textbook brought to St Gall by Marcellus, and asked for it to be included in a compilation of principally non-Bedan / non-Hrabanian computistica.

In this reading, an open discussion of this discipline appears to have taken place between the abbot and his schoolmaster in the years of Hartmut’s abbacy ( 872–883 ). That this was a time of intensive computistical debate at St Gall is evidenced by two original texts on this calendrical science written in 877 and 882, which remain extremely understudied ^[88]. The first was composed by an anonymous monk under the heading ‘Adbreviatio de pluribus compoti maioris necessitatibus’, and survives only in Cod. Sang. 459. The second, called ‘Interrogationes et responsiones de compoto Bedae’, was more popular, and the copies written for export name the author as Wichram, monk of St Gall. The ‘Adbreviatio’ shows some Irish features, which were also included by Wichram, who relied heavily on the earlier text. Notker Balbulus’ short note at the end of his Letter to Lantbert, urging the recipient to apply correctly the structure of the 19-year lunar cycle ( the sequence of common and embolismic years ) ^[89], is reminiscent of one chapter in both texts, headed De ciclo decennovenali et mensibus embolismis, which explains just that; one wonders if Notker himself could have been the author of the ‘Adbreviatio’. The ‘Adbreviatio’ was incorporated in a compilation that otherwise principally contains a calendar and a collection of computistical verse ( including those by Walahfrid Strabo ), while Wichram’s text was appended to a copy of Bede’s ‘De temporum ratione’ to which it was supposed to be a commentary. These two units ( calendar + ‘Adbreviatio’ + verses and ‘De temporum ratione’ + Wichram ) were written more or less at the time of composition of these two texts ( 877 and 882 ), and were soon thereafter bound together, with a full Easter table and its marginal ‘Annales Sangallenses brevissimi III’ additionally prefixed ( Cod. Sang. 459 ) ^[90]. Together this formed an impressive collection of Easter table, calendar, original St Gall texts, verses, and Bede’s ‘De temporum ratione’. Cod. Sang. 459 is therefore beautifully complimentary to Einsiedeln, Stiftsbibliothek, 321 ( 647 ), both being written in the decade of Hartmut’s abbacy ( 872–883 ).

This new focus on calendrical science under Hartmut translated more broadly into the ‘quadruvial’ or ‘physical’ arts right after his resignation, as impressively witnessed by the 645-page strong Cod. Sang. 250 of 889 ^[91]. The first 450 pages deal with computus, followed by the ‘Aratus Latinus’ under the rubric Astrologia and Hyginus, who was surely considered Astronomia. The computistical part contains an Easter table with the ‘Annales Sangallenses brevissimi II’ of 768–889 in the margin, Wandalbert of Prüm’s ‘Martyrologium’, a calendar with additional tables, and Bede’s three scientific works. More unusual are two shorter texts, the oldest surviving copy of the Irish pseudo-Columbanus ‘De saltu lunae’ and Ceolfrid’s paschal letter to the Pictish King Nechtan in a rare transmission separate from Bede’s ‘Historia ecclesiastica’. The main Bedan texts are closely related to Cod. Sang. 251, but here the insular abbreviations, source marks, and glosses are gone ^[92].

By the end of the ninth century, therefore, the St Gall holdings contained all of the essential computistical literature – and more – post ca. 700. What was missing was a collection of late antique texts on the subject. In the early Middle Ages, these circulated as a rather loosely defined corpus in what modern scholarship has termed the ‘Sirmond group of manuscripts’ ^[93]. The Reichenau had acquired such a collection in the early ninth century ( Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 586, then included in Paris, Bibliothèque nationale de France, Lat. 4860 of the late ninth century ^[94] ), while at St Gall it can only be found in a codex of ca. 900 ( Bremen, Staats- und Universitätsbibliothek, msc 0046 ). There is some overlap between the Reichenau and the St Gall collections, but the St Gall manuscript also contains some extremely rare texts, especially excerpts from a treatise on the Old Roman Easter reckoning abolished by Leo I in the mid-fifth century, the letter of the Spanish monk Leo of 627, and a Prologue to a Victorian Easter table of 699. The first two can also be found in the famous Hildebald Codex Cologne, Diözesan- und Dombibliothek, 83-II of 805, but the last is otherwise unattested. It has been demonstrated to be an Irish text, and occasional insular abbreviations throughout the manuscript suggest that the whole collection was copied from an insular exemplar ^[95]. It may be a stretch to argue that the collection as a whole was brought by Marcus and Marcellus to St Gall, and the relationship to the Reichenau codex Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 586 and the Cologne manuscript 83-II needs to be studied in detail. But the fact remains that this St Gall manuscript contains the only existing copy of another Irish text.

In sum, the arrival of Marcus and Marcellus at St Gall marked the beginning of intensified study of calendrical science in this monastery in the second half of the ninth century. The Irish element in this process is unmistakable, as represented by two unique Irish texts ( ‘Computus Einsidlensis’ and the ‘Victorian Prologue of 699’ ) and one for which the St Gall holdings transmit the oldest surviving full copy ( pseudo-Columbanus’ ‘De saltu lunae’ ). None of these received any wider reception ( ‘De saltu lunae’ only in the eleventh century from St Gall ). This suggests that St Gall had direct access to Irish calendrical texts that became available at the time of Marcus and Marcellus’ floruit at St Gall, and likely have been part of the books brought by them as related by Ekkehart.

For the present study, the ‘Computus Einsidlensis’ is of primary importance, because it is closely related to the ‘Munich Computus’ that applied the substitution of dies Mercurii with a neutral native term. The Einsiedeln Computus itself, however, does not contain a passage to that effect. Evidence that this was part of Marcellus’ thinking, and presumably his teaching as schoolmaster, comes from the best documented of his activities at St Gall, charter writing.

Four St Gall charters were written by a certain Marcellus, all surviving as originals ^[96]. Some doubt has been expressed in identifying this person with the Irish schoolmaster, principally on account that Marcellus is a common enough name and that the charters were written in Continental script ^[97]. Neither argument is convincing. In the St Gall ‘Necrologium’, the name only appears once, and likewise in the St Gall lists in the Reichenau ‘Liber confraternitatum’ and the Pfäfers ‘Liber viventium’ ^[98]. As for script, the learning of a new script is not an impossible task and polygraphism of scribes in the early and high Middle Ages was not unheard of ^[99]; here, it may be enough to point to the Irish teacher Martin at the Cathedral school of Laon at roughly the same time, who exclusively wrote in Caroline minuscule ^[100].

Marcellus’ charters are often dated to 853–860, but their dating criteria are more easily aligned with 847–853 ^[101]. In the second of his charters ( St Gall III 210 ), Marcellus added a unique remark on the weekday descriptor for Monday: “day of the moon, which Christianity better calls second feria” ( diem lunae, quam Christianitas melius secundam feriam vocat ). This very much echoes Augustine’s comment on the naming of Wednesday after a pagan God: “an ecclesiastical manner of speaking certainly comes better out of a Christian mouth” ( melius ergo de ore christiano ritus loquendi ecclesiasticus procedit ). Classroom teaching translated here into charter writing. The Irish schoolmaster initiated a fundamental change in dating terminology in St Gall charter writing, as from the 850s/860s the use of Roman planetary weekdays turned from dominant to occasional practice. His last charter suggests that Marcellus may not have applied this practice systematically himself, but his students did, especially Notker Balbulus, whose numerous charters all use the feria terminology. Borgolte is very much to the point in his summary of this process ^[102]: “Der irische Mönch hat also nicht konsequent auf eine Datierungsreform hingesteuert, sondern in einer von ihm geschriebenen Urkunde und sicher auch in der mündlichen Unterweisung seiner Schüler den nichtchristlichen Ursprung der gebräuchlichen Wochentagsnamen ins Bewußtsein gehoben.”

Notker Labeo and Fáelán

The impact of St Gall’s vibrant intellectual endeavour of the second half of the ninth century was still felt a century later. St Gall scholarship reached a new height under Notker Labeo ( †1022 ), who is especially known for his unprecedented translations from Latin into Old High German of Biblical and mid- to higher educational texts. In a letter to Bishop Hugo of Sion, written towards the end of Notker’s life, the St Gall schoolmaster gives a detailed account of his translation activity ^[103]:

Sunt enim ecclesiastici libri – et pręcipue quidem in scolis legendi –, quos impossibile est sine illis pręlibatis ad intellectum integrum duci. Ad quos, dum accessvm habere nostros uellem scolasticos, ausus svm facere rem pene inusitatam, ut latine scripta in nostram [ linguam ] conatus sim uetere, et syllogystice aut figuarate aut suasorie dicta per Aristotelem uel Ciceronem uel alium artigraphum elucidare.

Quod dum agerem in duobus libris Boetii, – qui est ‘De consolatione Philosophiae’ et in aliquantis ‘De sancta trinitate’, – rogatus [ sum ] et metrice quedam scripta in hanc eandam linguam traducere, ‘Catonem’ scilicet ut ‘Bucolica’ Vergilii et ‘Andriam’ Terentii. Mox et prosam et artes temptare me uoluerunt, et transtuli ‘Nuptias Philologię’ et ‘Cathegorias’ Aristotilis et ‘Pergermenias’ et principia arithmetice. Hinc reuersus ad diuina totvm psalterivm et interpretando et secundum Augustinum exponendo consummaui; ‘Iob’ quoque incepi, licet uix tertiam partem exegerim. – Nec solvm hec sed et nouam rhetoricam et compvtum nouum et alia quedam opuscula latine conscripsi.

For there are ecclesiastical books – and particularly those to be read in the schools – that are impossible to be led to a full understanding without those [ books on the arts ] having been enjoyed before. As I wished our students to have access to these, I dared to do something almost unprecedented in that I tried to turn texts in Latin into our language, and to elucidate something expressed syllogistically or figuratively or exhortatively through Aristotle or Cicero or other authors of the arts.

While doing this for two books of Boethius – who is ( known ) for the ‘On the Consolation of Philosophy’ and the considerable ‘On the Holy Trinity’ –, I was asked to translate certain works in metre into the very same language, namely the ‘Cato’ as well as Vergil’s ‘Bucolica’ and Terentius’ ‘Andria’. Soon, they wanted me to try prose and the arts, and I translated ‘The Marriage of Philosophy’, Aristotle’s ‘Categories’ and ‘Peri Hermeneias’ and the principles of arithmetic. Then I returned to the divine, and completed the entire psalter by both translating and explaining according to Augustine; I also began ‘Job’, even if I have hardly completed a third of it. – Not only these, but I composed a new rhetoric, a new computus, and some other little works in Latin.

For the purposes of the present study, the references to the computus and the psalter are of particular interest. The ‘computus novus’ – the only of Notker’s original works that survive in both Latin and Old High German – is more appropriately entitled “On four questions of computus” ( ‘De quatuor quaestionibus compoti’ ) after its heading, with one of its witnesses ( Paris, Bibliothèque nationale de France, nal 229, fols. 10v–14v – Weissenburg?, ca. 1100 ) identifying not only Notker as author ( as do other copies written for export ), but also his pupil Ekkehart, the author of the ‘Casus sancti Galli’ cited above, as recipient ^[104]. Notker’s intention was not to write a textbook on this discipline for the St Gall school, but a commentary to Helperich’s Computus of 903, which seems to have arrived in the Lake Constance region only in the second half of the tenth century ^[105]. The four questions raised by Notker concern the principles of Easter at the beginning of his work and the intercalated day ( bissextus ) at the end, while the core – the predominant reason for writing this commentary – focusses on the structure of the 19-year lunar cycle and the length of the synodic lunar month ( the period from one new or full moon to the next ). For these two central themes, Notker and his students found in the St Gall library holdings better and much more detailed explanations than those provided by Helperich. The pseudo-Columbanus tract ‘De saltu lunae’ calculates the mean synodic lunar month as 29 days, 12 ½ hours, and almost 10 moments. It appears that Notker was challenged by his students who argued that this was only an approximation, and that an exact value can surely be calculated. Notker had none of this, and in true schoolmaster fashion reverted back to authority, declaring that those who would consider themselves wiser than Columbanus may try to arrive at a more precise result. Ekkehart’s contemporary Hermann of Reichenau accepted this challenge in the 1040s, which famously led him to the construction of a new lunar calendar based on exact data ^[106].

This episode shows that an Irish text that – uniquely – entered the St Gall library holdings and curriculum at the time of Marcellus and his students in the mid- to late ninth century was still taught at St Gall in the late tenth and early eleventh centuries and initiated one of the best documented scientific debates of that time. It will not be a coincidence that during Notker’s formative years an Irish schoolmaster, Fáelán, was active at St Gall, who died on 3 June 991 according to the necrology and the ‘Annales Sangallenses Maiores’ ^[107]. In the necrology, he is remembered as a “most learned and most kind teacher” ( doctissimus et benignissimus magister ) and, significantly, Hermann of Reichenau was aware of Fáelán’s reputation, calling him “erudite Irishman” ( Scottus eruditus ) in his Chronicle ^[108]. Unfortunately, Ekkehart’s narrative in the ‘Casus sancti Galli’ breaks off at 972, and therefore does not mention Fáelán, whose floruit at St Gall coincides with a renewed interest in calendrical science, with a decidedly Irish flavour.

Two copies of ‘De saltu lunae’ were made at St Gall in the 970s or 980s, one of which was incorporated – incomplete – in Cod. Sang. 459 ( pp. 125–126 ). The other is Zurich, Zentralbibliothek, C 62, fol. 127r–v, which is related to another St Gall codex of the last third of the tenth century: Vatican, Biblioteca Apostolica Vaticana, Vat. lat. 644 ^[109]. In its tabular material, Vat lat. 644 is based strongly on the Karlsruhe Bede ( Karlsruhe, Badische Landesbibliothek, Aug. perg. 167 ) of the 840s ^[110]. It is important that the scribe of Vat. lat. 644 evidently worked from Aug. perg. 167 in an Irish hand, and not from its sister codex Cod. Sang. 248 in Continental script ^[111]. Evidently, the scribe of Vat. lat. 644 had no problems reading Irish script, and may even have preferred it. By copying the feastdays of the three Irish national saints Brigit ( 1 February ), Patrick ( 17 March ), and Columba ( 9 June ) and also St Gall ( 16 Oct ), and adding Abbot Otmar of St Gall ( 16 November ) and Pirmin of Reichenau ( 3 November ), the calendar of Vat. lat. 644 tells the story of copying an Irish script book at the Reichenau (?) for St Gall consumption.

Some of the tables of the contemporary St Gall manuscript Zurich, Zentralbibliothek, C 62 that contains pseudo-Columbanus ‘De saltu lunae’ are also directly copied from Aug. perg. 167, here with some more textual elements ^[112]. The script of C 62 is closer to St Gall style than Vat. lat. 644, so there are at least two scribes at work in the later tenth century turning Irish script tables and texts into Caroline minuscule. The scribe of C 62 also revises the material copied, including ‘De saltu lunae’, demonstrating that the copyist was a well-trained computist.

From this evidence, it appears that Fáelán was eager to include Irish material in his St Gall teaching, and this may well have extended to other essential calendrical concepts, like Augustine’s warning not to use Roman planetary weekdays and ideally to replace them with more appropriate native terms. This will have influenced Notker, who, in his work on the psalter and probably led by Fáelán, likewise drew on manuscripts and expertise of the time of Marcus and Marcellus. Notker’s Letter to Hugo mentions only two Biblical books, Job and the psalter, and for the psalter also Augustine’s commentary. The Augustine commentary he used is evidently the six-volume edition of the 850s / 860s mentioned above, and his Old High German discussion of Psalm 93 is based almost exclusively on it ^[113]. Notker starts by reproducing the psalm’s Latin title, Psalmus ipsi David quarta sabbati, which is then rendered into Psalmus Christo in mittauuechun ^[114]. Commentators on this passage have frequently invoked Augustine’s commentary as the underlying source for Notker here. But by quoting only Augustine’s phrase quarta feria, qui Mercurii dies dicitur a paganis, et a multis Christianis and therewith simply stressing that Notker’s mittauuechun is a translation of Augustine’s dies Mercurii, they have somewhat missed the point ^[115]. The crucial part in Augustine’s commentary is the immediately following sentence, urging Christians to use their own language to replace such a reference to a pagan God. If read literally, as the Irish author of the ‘Munich Computus’ did in 718/719, this applied only to the topic of this psalm, the fourth day of the week, Wednesday. This solves the historian’s and linguist’s puzzle of why only the pagan reference of this one weekday was substituted with a more neutral term ^[116]. It has nothing to do with the pagan god referenced, Wodan, but with the literary context of its first attested occurrence, a literal application of Augustine’s advice given in his commentary on Psalm 93. In this, Notker followed an old Irish tradition introduced into St Gall’s intellectual tapestry by Marcellus in the mid-ninth century.

Conclusion: Irish Influence at St Gall

The question of Irish influence on the intellectual culture of early medieval Europe has been widely discussed over the past two centuries, and here is not the place nor space to outline the debate in any meaningful detail. The views range, in the words of the St Gall librarian Johannes Duft in a famous article of 1955, from ‘Iromanie’ to ‘Irophobie’. Duft himself engaged in the debate more than once, principally from a St Gall centred perspective. He rightly warned against the use of the term ‘Irish colony’ for St Gall ( and implicitly other Continental monasteries ) ^[117]. It is indeed difficult to prove institutionalised and continuous Irish presence over centuries, let alone that any of these monasteries would have been dominated by Irish monks, if any of this is the definition of a ‘colony’. But he also argued that individual Irish intellectuals asserted some sporadic influence, though he overstressed that most of their knowledge was acquired on the Continent rather than imported from Ireland ^[118]. For Irish texts, he rightly observed that these can mostly be found in Continental copies rather than in Irish script ( which, it will be noted, implicitly suggests a meaningful number of literati on the Continent being able to read and convert Irish into Continental script ) ^[119].

In his recent ‘The Irish Scholarly Presence at St. Gall’, Sven Meeder developed many of Duft’s arguments considerably further. First, Meeder shared Duft’s assessment that books in Irish script were widely considered useless by a Continental monastic community and if kept, this happened for not much more than ornamental purposes ^[120]. This may be true for some of the literati of a given community, but it does not take enough into account certain areas that demonstrate substantial engagement with Irish script books, like Continental glosses in them or the vibrant copying activity from Irish into Continental script that required that copyists at least did not have much difficulty reading Irish script. Much more work needs to be done in such areas to fully appreciate the Continental engagement with Irish script books in the early Middle Ages. A good example is the copying from Irish into Continental script of large parts of the Karlsruhe Bede ( Aug. perg. 167 ) of the 840s by two scribes of the second half of the tenth century for consumption at St Gall discussed above. One of the copyists may have been an Irishman, the St Gall schoolmaster Fáelán. If so, this is another interesting example of Irish scholars having no problems writing in Continental script.

Second, Meeder argued that Irish monastic foundations on the Continent received their Irish written knowledge ( in form of texts ) principally through Continental channels rather than directly from Ireland. This may well be the case, but it will be important to analyse further how ‘Irish’ these ‘Continental’ channels and networks may have been. Often we can date the acquisition of scholarly works and determine their Continental movement, and we need to understand better the role played by personal networks of known Irish scholars working on the Continent. For St Gall and the wider Lake Constance area, e. g. the links to the circle of Sedulius Scottus in Liège – discussed by Meeder ^[121] – and of Martin of Laon seem to have been instrumental in the distribution of knowledge, be it Irish, English, or Continental. Martin’s example – he wrote exclusively in Caroline minuscule – is also a reminder that script probably is not a good indicator in determining Irish involvement in the transmission of knowledge, as one would expect Irish scribes and scholars working in a Continental setting to apply Irish script for private use, Continental script for wider consumption in a given monastic community – in the same way as English is used today by non-native speakers, largely without major difficulties, to reach a wider audience or to fulfil the obligations of their employers. Irish abbreviations in Continental script books may be more indicative of Irish involvement at some stage in the transmission process, and much more wide-spread; when studied carefully, they will reveal, e. g., if a text or a book was directly converted from an Irish script exemplar into Caroline minuscule, as is the case for the ‘Computus Einsidlensis’ in this study.

The ‘Computus Einsidlensis’ raised doubt about Meeder’s third main argument, that no scholarly work of Irish origin is known to have been brought to St Gall directly from Ireland, particularly by Marcus and Marcellus ^[122]. He decided to work from selected examples ( ‘De duodecim abusivis’, the ‘Collectio canonum Hibernensis’, Ailerán’s ‘Interpretatio’, and Finian’s and Cummian’s penitentials ). The analysis of the transmission of these texts led him to the conclusion that St Gall received these works not from Ireland, but from other Continental centres. It is dangerous to project this conclusion to the entire St Gall holding, though, and to treat St Gall as representative of other centres with Irish connections. In terms of book acquisition – a costly endeavour – St Gall was a latecomer, because its finances were squeezed by the bishop of Constance until the 820s ^[123]. When St Gall had the economic means to stock a library, Irish texts had already widely spread in Europe, and it was only natural to acquire these from the closest libraries and through established channels ( of Irish connections? ), for Irish texts from the Reichenau, Bobbio, north-eastern France, and other places. A more systematic analysis of the St Gall holdings would surely change the picture painted by Meeder ^[124]. Certainly, the computistical evidence underlying the present study leads to different results. The St Gall library held three Irish texts of this monastic discipline. Two of these, the ‘Computus Einsidlensis’ of ca. 700 and the ‘Victorian Prologue of 699’, are unique witnesses, which in itself suggests that they were probably brought directly from Ireland, certainly in combination with the fact that they left no traces elsewhere. The abbreviations in the ‘Computus Einsidlensis’, as mentioned above, prove that it was copied from an insular exemplar, and the context of copying makes it highly likely that this text was in the travel library of Marcus and Marcellus when they arrived at St Gall in the 840s.

The transmission history of the third text, pseudo-Columbanus’ ‘De saltu lunae’, is more complex, and leads to Meeder’s fourth main argument, that Irish foundations on the Continent played no major role in the further dissemination of Irish texts. A short version of the core idea of this work circulated widely in the eighth and ninth centuries, and the full version left few but regular traces ^[125]. Still, the oldest surviving copy of the full text is a St Gall manuscript of 889, and until Notker Labeo started an influential discussion of its content in ca. 1015, the full text is not attested elsewhere. Afterwards, it radiated out from St Gall into southern Germany, often in combination with Notker’s ‘Quaestiones’; therefore, St Gall did distribute Irish knowledge.

‘De saltu lunae’ is a classic example of how Irish ideas influenced St Gall’s intellectual endeavour, and that of the Lake Constance region more generally. The ascription to Columbanus will have identified this text as Irish to the St Gall community, but this was not the reason for it to be studied. In the discussion of Irish intellectual influence on the Continent, a lot of weight is frequently put on the question if the recipient culture knew that a concept or text was Irish. From the perspective of an historian of ideas, this does not appear to be a very helpful category. Bad ideas need authority, good ones do not. Physicists use the theory of relativity because it is a good idea, not because it is from Einstein. Pseudo-Columbanus had something unique to offer to the computistical thought-world of the early Middle Ages, and Notker, as well as, importantly, very likely his Irish teacher, Fáelán, understood this, as did Hermann of Reichenau a few decades later.

A noticeable scholarly tradition that started with the arrival of Marcus and Marcellus in the mid-ninth century can therefore be reconstructed on various levels, and it helped initiating St Gall’s ‘Golden Age’ of the second half of the ninth century. Some of this tradition can be established through the library holdings, while ideas transmitted through oral teaching are more difficult to grasp. One prime example is the application of Augustine’s warning against using attributions to pagan Gods for weekdays in general, for the fourth day of the week, Wednesday, in particular. Marcellus introduced this Irish tradition into St Gall teaching, and it ultimately led, possibly through Fáelán, to Notker substituting German ‘Wodan’s day’ with ‘middle of the week’. It is no coincidence that the renaming of Wednesday into ‘Mittwoch’ in Old High German written culture happened in a centre with Irish connections, without which this would not have happened.