This brief article undertakes to alert the academic community to the existence and digital availability of a largely unknown Byzantine lectionary and to establish, contrary to the hypothesis put forward a few years ago, that it was destined for public, not private, use. In the process, as a result of direct examination of the manuscript in the last two years, for the first time ever this article describes aspects of the lectionary's content, identifying the scriptural loci of certain pericopes and their liturgical use. It also indicates the exact structure and length of its sections. In so doing, the article corrects misrepresentations of the lectionary's content in the two relevant studies published so far, by Albrecht Gerber and Jean-Marie Olivier, in the journal of the Australian Institute of Archaeology, Buried History (2017, 2019); these works are engaged throughout. After introducing the manuscript's current location and library description, the article summarises the discussion of Gerber and Olivier in regard to its provenance, and then tackles what the two researchers reported on its content and possible use. The article ends with a tentative sketch of the community that might have used this lectionary.

The object catalogued as ‘Additional Manuscript’ (or Add. MS) 40 in the University of Sydney Library's Rare Books and Special Collections is a very little-studied item. Technically, this is a Byzantine liturgical manuscript, a Gospel lectionary, written on parchment leaves, recto and verso, in two columns and in minuscule script. It is commonly known as the Codex Angus, where ‘codex’ obviously refers to its book format – typical for scriptural manuscripts from the dawn of Christian literary cultureFootnote ¹ – while ‘Angus’ denotes the name of its onetime owner, Samuel Angus (1881–1943).Footnote ² The manuscript was deposited in the library in 1936, immediately after its acquisition from Adolf Deissmann, Angus’ Berliner mentor.Footnote ³ The catalogue description indicates Constantinople as its probable place of manufacture, which, given its poor workmanship, numerous mistakes and overall appearance (see Figure 1),Footnote ⁴ is doubtful. (This is not to say, however, that the Codex Angus did not belong to the liturgical sphere of Constantinople. Daniel Galadza's comparative study of lectionaries from Constantinople and Jerusalem brings to light similarities and differences that help in locating the manuscript under consideration in the Constantinopolitan tradition.Footnote ⁵) The same description refers furthermore to the ‘Greek Church in Bulgaria’ as place of provenance. Various dates of production have been put forward, from the eleventh to the fourteenth century, with the latter date being ascribed by Basilis Katsaros's list of parchment manuscripts that belonged to Kosinitsa Monastery before its library was robbed during the First World War.Footnote ⁶ Steps towards carbon dating the parchment leaves are currently underway.Footnote ⁷ The manuscript was digitised in 2017, before the restoration of its binding in 2021, with funds from the Selby Old Foundation.

Figure 1. Codex Angus, fos 31v, 29r: examples of the poor quality of materials and workmanship.

Except for catalogue entries, until recently the codex was largely forgotten. Only Gerber and Olivier have so far discussed it at any length. Neither of them examined the manuscript's content thoroughly, however, instead focusing upon its possible provenance and the historical vagaries of its journey from Greece to Bulgaria to Germany to Australia.Footnote ⁸ It is noteworthy that, while Gerber was of the view that from Constantinople the manuscript reached Bulgaria ‘many centuries ago’, where it was supposedly kept in monasteries and museums,Footnote ⁹ Olivier prudently concluded that it belonged to Kosinitsa Monastery, in Northern Greece, which Bulgarian raiders ransacked in 1917.Footnote ¹⁰ Thus, the codex reached Bulgaria at the end of the second decade of the twentieth century, not earlier. Assuming the conclusion of Olivier to be correct, and if item 58 in Katsaros's list is what currently is known as the Codex Angus, the catalogue of the University of Sydney Library should record Kosinitsa Monastery as its place of provenance, not the ‘Greek Church of Bulgaria’. That said, except for Olivier's correction about the Bulgarian leg of the journey, Gerber's account remains a credible source for the modern transition of the lectionary through various time zones, from Bulgaria to Germany to Australia.

More relevant, here, are the points made by Olivier and Gerber on the content of this manuscript. Olivier's article has precious little to say in this regard and the information it supplies is not entirely free of error. According to him, this codex would be a New Testament lectionary that contains ‘readings for the days of the week between Easter and Pentecost and Saturdays/Sundays of the other weeks’.Footnote ¹¹ In reality, it is a Gospel lectionary that comprises 220 unnumbered passages, so counted by this writer. It does not include other New Testament pericopes and has most of the usual pericopes for the Easter to Pentecost season, including for weekends, as well as weekdays.Footnote ¹² Olivier is correct when he points out that the codex has the weekend readings for the rest of year, namely, the other two liturgical seasons of the Byzantine rite, that is, the weeks after Pentecost and the Lenten section, the latter together with Holy Week.Footnote ¹³ Nevertheless, for Holy Week it prescribes readings for each day. But the manuscript includes two more categories of passages, which Olivier does not mention, that is, for the festivals and the saints of the twelve months from September to August (the menologion section, in technical parlance),Footnote ¹⁴ as well as the eleven Gospel readings for Sunday matins.Footnote ¹⁵

In turn, Gerber's article includes a longer description of the manuscript,Footnote ¹⁶ which in parts requires correction too. This description refers to the codex's wooden cover, measurements, scribal competence and the overall appearance of the parchment leaves, especially its lack of illuminations. Indeed, the only elaborated adornment is the floral depiction in red ink that serves as a header for the first folio (see Figure 2).Footnote ¹⁷ The artistic quality of this ornament does not impress, however. Several of Gerber's assertions are problematic, and in what follows I dispute one of them in particular. Before that, another two of his points are worth addressing, for the purpose of forming a more accurate idea of the codex.

Figure 2. Codex Angus, fo. 1: the most elaborate illustration in the entire manuscript.

Following the pencilled page numbers, undoubtedly added by a modern hand, Gerber gives the total as 122 parchment leaves. But this count is not accurate, as one folio is missing and Gerber did not realise this. By all accounts, it was already missing when the codex was digitised, in 2017, the year when his own article was published, and its absence went again unnoticed. Thus, the digital version does not include the fifteenth folio; nor does the original, unfortunately. This leaf must have been lost at some point between 1936 – when the manuscript reached the library and when, presumably, the leaves were numbered in pencil – and digitisation. The missing text is obvious in that the last pericope on fo. 14v is incomplete (John viii.51–9; Tuesday of the fifth week after Easter), whereas the first reading on fo. 16 represents the end of another pericope (John x.9; Thursday of the fifth week after Easter). What matters, for the record, is that the Codex Angus has now only 121 parchment leaves. The other matter of interest is Gerber's point that, after the Easter to Pentecost section, the lectionary contains readings ‘for each week (Sunday) of the year’.Footnote ¹⁸ In fact passages are prescribed for Saturdays too, hence the reference to the section dedicated to the weeks after Pentecost as σαββατοκυριακόν (Saturdays and Sundays),Footnote ¹⁹ with Holy Week having prescribed daily readings.Footnote ²⁰

More relevant here is another of Gerber's assertions, namely, that the codex was destined for private use within a monastic setting. He mentions private use twice, monasticism being referred to in the second instance. Gerber's first statement is not very explicit. In his words, ‘this lectionary was clearly not intended for display’.Footnote ²¹ Poor quality of materials and the unaesthetic appearance are taken to signify the codex’s deliberate production for anything but liturgical use. One wonders how warranted this conclusion can be. If, as will be seen, the manuscript belonged to a small community of modest means that needed it for simple worship, not cathedral pomp, poor quality and appearance cannot signify a private destination and use. But the context leaves no doubt about Gerber's assumption that, because of its appearance, the codex could not have been meant ‘for display’. It is this assumption that led him to the conclusion that it was not designed for public worship.

This conclusion becomes obvious several paragraphs later, when Gerber returns to this point and adds the following: ‘The original purpose of this lectionary was primarily for personal, that is to say, devotional use, as a kind of “work copy” for cenobitic monks or even consecrated anchorites.’Footnote ²² That he means private use is clear, given the known practice of monastic meditation on sacred texts. The spiritual discipline of monastic meditation on scriptural passages is abundantly attested.Footnote ²³ Nevertheless, Gerber does not supply evidence for either the supposed monastic setting where this lectionary would have been used or the reasons that demanded the private use of a lectionary, whether by ‘consecrated anchorites’ or others. Olivier's discovery that the codex was stolen from a Greek monasteryFootnote ²⁴ lends support to the idea of monastic use, but Gerber could not have been aware of this information when he published his article. Even so, what Olivier's findings indicate is the presence of this manuscript at Kosinitsa Monastery for an unspecified time, not that it was manufactured thereFootnote ²⁵ and destined for private use.

The hypothesis that a lectionary could have been destined for any other purpose than public worship finds no support among scholars. As a medieval invention,Footnote ²⁶ lectionaries, Jeffrey Anderson tells us, were specifically designed for public use.Footnote ²⁷ A plethora of scholars, of various generations, agree with this assessment.Footnote ²⁸ Roland Betancourt, likewise, emphasises that for the Byzantines reading the Scriptures was foremost a public, liturgical act, his conclusion drawing upon the study of numerous lectionaries.Footnote ²⁹ After examining a great many manuscripts, one of Gerber's sources, Christopher Jordan, arrived at the same conclusion. In his words, ‘It seems that Gospel lectionaries are liturgical codices designed for public recitation rather than private reading. Any Gospel reading by monks outside of the liturgical services probably involved [full] Gospel manuscripts, the non-liturgical Gospels.’Footnote ³⁰ Gerber did not follow Jordan's analysis up to this conclusion. But let us consider Gerber's argument more closely.

Gerber invokes three reasons for concluding that the codex was not meant for the liturgical setting. First, he mentions the minimal ornamentation, the clumsy calligraphy and the low literacy levels of certain contributing scribes, obvious in the many ‘orthographical errors, erasures, overwrites or corrections’ that litter the text.Footnote ³¹ For him, substandard workmanship would be incompatible with public worship. Second, Gerber observes that ‘the text lacks any ekphonetic notations – a system of mnemonic voice-modulation marks, usually in red ink, to guide the (audible) reader’s intonation, tempo or pitch’.Footnote ³² His assessment relies on Jordan's doctoral dissertation on Byzantine lectionaries, adopting the quantitative argument the author put forward there.Footnote ³³ In summary, this argument posits that, since many Byzantine lectionaries contain ekphonetic notation, denoting their use for public worship, the absence of this feature means private use. Third, to assert the codex’s private or devotional usage, Gerber refers to an eleventh-century list of manuscripts that mentions one such ‘work copy’.Footnote ³⁴ This external proof is unconvincing, as it cannot be verified and lacks quantitative support. That said, even if someone would amass quantitative evidence to that effect, the internal evidence tells a different story. It is that evidence that is the contribution of the present article. Gerber himself does not delve into this topic, shifting his attention to a related point, that ‘exactly when, where and for what type of monastery the Codex Angus was initially produced remains undetermined’.Footnote ³⁵ While this point has no bearing on the matter of whether the manuscript was used privately or publicly, it draws attention to the issue of manufacture and destination.

It is Gerber's view that the codex was meant for private, not liturgical, employment, monastic or otherwise, that is challenged here. The Codex Angus is a liturgical object, designed for public use. The absence of ekphonetic notation does not necessarily indicate a private destination.Footnote ³⁶ After all, as Betancourt shows, not all lectionaries were equipped with ‘chant marks’, despite being used in worship gatherings.Footnote ³⁷ That intonation signs are missing from the Codex Angus is consistent with the poor workmanship of this manuscript, but this is a different matter altogether. It cannot be invoked in support of Gerber's hypothesis. Further evidence in this regard renders his conclusion improbable.

Gerber ignores the fact that the lectionary's inferior manufacture denotes yet another possibility, which has no bearing on the matter of usage. In short, it relates to the financial capacity, or lack thereof, of the community that owned and used the manuscript.Footnote ³⁸ Parchment was an expensive material,Footnote ³⁹ and even unskilfully written books increased the cost of the final product. In addition, the price range of the best scriptoria, which produced exquisitely illustrated and elegantly written tomes, remained prohibitive for disadvantaged communities. It is reasonable to think that, except for cathedrals, rich urban parishes and large monasteries, countryside churches and small monastic settlements could not afford the best liturgical items that the scriptoria of the major centres produced. This should not come as a surprise. Throughout history, luxury items, including of a religious nature, have been and are available only to the affluent. And the codex under consideration does not exhibit the known features of a luxury item destined for, say, cathedral ceremonies,Footnote ⁴⁰ such as illuminations, elegantly stylised, gold letteringFootnote ⁴¹ and ekphonetic notations. Its appearance matches the minimally adorned tomes of the four Gospels in their entirety, as well as, immediately relevant to this article, a group of lectionaries designed for reading during services, but not for processions.Footnote ⁴² It is possible that these unadorned lectionaries were minimalistic in their outlook precisely because of the lack of means of the communities that needed them. Such is, undoubtedly, the case of the community that owned the Codex Angus, which could not have paid for the services of the accomplished hands at major scriptoria. No wonder that there are other pointers to its poor manufacture, for example the collage of leaves written by different scribes, with the text being inexpertly squeezed so that continuity is secured. This is the case of fos 27v and 28r (see Figure 3).

Figure 3. Codex Angus, fos 27v, 28r: attempts by the scribes to secure the continuity of the text.

In short, the poor quality of the parchment, the minimal adornment and the many inexpert hands that copied the text denote its manufacture by a ‘budget’ scriptorium, which produced liturgical books for small communities, either monastic or parochial, that could not afford a quality item.Footnote ⁴³ The large and wealthy Kosinitsa Monastery – if the Codex Angus is the mysterious object catalogued there – does not fit this profile. This might mean that the manuscript was deposited there at some point out of respect for its liturgical character, after being used elsewhere,Footnote ⁴⁴ not that it was used privately.

The time has now come to turn to Gerber's points about usage. Amounting to a selection of scriptural passages prescribed for liturgical reading, similar to any other Gospel lectionary, the Codex Angus could not have been destined for private use. Private readers of Scripture, including for the purposes of monastic contemplation and prayerful reflection, would seek manuscripts of the New Testament, or the collected Gospels in their entirety, not lectionaries that provide a very limited array of passages. Jordan's conclusion to that effect confirms this point. In this light, since the Codex Angus is a lectionary, not a New Testament manuscript, it was destined for public worship not private reflection. Abundant internal evidence supports this assessment: first, clear marks of heavy usage and, second, aspects of its content.

The most important proof of the codex’s public use comes from its overall appearance, which does not corroborate the idea of reflective reading in the privacy of one's monastic cell or home. The evidence of public use is overwhelming. For example, the front and back leaves are weathered and very dark (see Figures 2 and 4), which shows that originally – and for an indeterminate length of time – the lectionary did not have the current wooden hard covers to protect it. While this evidence supports the proposal, that the Codex Angus originated in a ‘budget’ scriptorium that manufactured liturgical books for disadvantaged communities, it also substantiates the assertion that it was heavily used for liturgical purposes.

Figure 4. Codex Angus, fo.122v: the last leaf, showing clear signs of exposure to the elements.

Indeed, throughout the manuscript can be found many stains left by a variety of liquids, such as darker ones, of wine accidentally poured upon it during eucharistic celebrations, but also lighter, left by water, possibly spilled on the occasion of baptismal offices and the blessing of the waters, as well as oil.Footnote ⁴⁵ Interestingly, fo. 116 appears to have been sprinkled with scented oil, to the extent that when one opens it a very pleasant aroma can be sensed. Other stains are caused by melted wax from candles (see Figure 3), consistent with the rush of reading the designated passages during liturgical offices.Footnote ⁴⁶ More often than anything else, there are stains left by greasy fingers and unwashed hands. As expected, these appear on the lower outer corners of the pages, both recto and verso, as well as on the median outer side of the leaves (see Figure 5).Footnote ⁴⁷ Also, not using anything like the Jewish ‘Torah’ pointers (the so-called Yad), but using one's own fingers to follow the text, means that the script is faded or indeed vanishing in many places. Sometimes, the fading text can still be read,Footnote ⁴⁸ but where repeated pressure led to its disappearance later scribes added the missing text in their own handwriting (see Figure 6).Footnote ⁴⁹ All this is indication of a liturgical book heavily used for public worship.

Figure 5. Codex Angus, fo. 69r: deterioration caused by touch.

Figure 6. Codex Angus, fos 91v, 5r: heavily stained parchment and repairs to deteriorated text.

Second, regarding the content of the manuscript, noteworthy is the very ‘seasonal’ aspect of its selection of passages. Regular lectionaries, such as the Codex Atheniensis 2803 (dated 1288),Footnote ⁵⁰ give pericopes for the entire liturgical year. The Codex Angus, however, corresponding to Dumbarton Oaks ms 1, BZ.1939.12,Footnote ⁵¹ gives a short selection of passages, especially for the longest section of the liturgical year, that is, the thirty-six weeks after Pentecost (from fo. 25r to fo. 52v). In total, there are only twenty-eight leaves for this section, compared to the eight weeks of the Paschal season, which are allocated a total of twenty-four and a half leaves (from fo. 1r to fo. 25r), and the seven weeks of the Lenten season, including Holy Week, which occupy forty leaves (from fo. 56v to fo. 96v). More specifically, with utmost consistency readings for the largest section of the year are prescribed – and marked as such by paratextual numbered notationsFootnote ⁵² – only for Saturdays and Sundays, hence the name σαββατοκυριακόν. As this part of the liturgical year corresponds to the northern hemisphere's summer, autumn and winter, it includes seasons of busy agricultural and farming activities, as well as difficult times for travelling to church. Other manuscripts give the same limited selection of readings for the σαββατοκυριακόν. Nevertheless, while a study of the socio-economic factors that might have determined this reduction in terms of prescribed readings is not usually undertaken, in the case of the Codex Angus this feature seems to denote the rhythms of a small community, very possibly a rural one of farmers and shepherds,Footnote ⁵³ but also one usually badly affected by the snow in winter. That winters could have been hard in that area is obvious in the very few readings that the Codex Angus prescribes for Christmas (25 December): only for vespers on the eve and for the festal liturgy (not even for matins, let alone the second and the third days of the festival).Footnote ⁵⁴ It is true that the rubrics point to other places within the lectionary where the passages can be retrieved, but the fact that they are not copied here is telling.

In this last regard, my argument is based on the practical rationale of lectionaries as liturgical books, where the presence of readings indicates compulsory gatherings for worship.Footnote ⁵⁵ If there are no readings, no gatherings are held. This is another important proof that the manuscript under consideration was meant for public worship. In short, the scarcity of prescribed readings was very likely determined by the geographic features and the economic rhythms of the community that used it. As such, this manuscript documents the fluctuations of liturgical practice along the lines of Galadza's conclusions to that effect.Footnote ⁵⁶ This and the other specifics point to the Codex Angus as a liturgical item meant for public worship and used as such by a small community. Thus – before it was stored at Kosinitsa Monastery, if Olivier's conclusions are correct and if this is the item Katsaros catalogued – this lectionary belonged to a rural community, possibly located in Northern Greece's mountainous regions, near Kosinitsa Monastery. The Codex Angus, which does not evidence the rigorous observance of monastic liturgical standards, thus gives indirect insights into a small Byzantine community as it navigated everyday life and religious commitments.