Early medieval vernacular Celtic glosses: originals or translations? A case study on the Vienna Bede

This study investigates the Old Irish glossing tradition on the Venerable Bede’s De Temporum Ratione, a computistical work from the early eighth century. Its main source is the Vienna Bede, a fragmentary manuscript with Old Irish and Latin glosses dating from the late eighth/early ninth centuries. It focuses on parallel glosses found in the Gloss-ViBe corpus where the Vienna Bede has an Old Irish gloss and the other manuscripts feature glosses in another language (Latin or Old Breton/Welsh). Minute analysis of individual glosses is used to determine whether early medieval vernacular Celtic glosses originals or translations from Latin glosses? The heterogenic nature of early medieval gloss corpora makes this a complex question for which there is no straightforward answer: for some glosses, a translation from Latin into Irish is almost inevitable, but others suggest Irish influence on the Latin parallel glosses. Accordingly, each case is discussed individually and the results are synthesised in the final part of the article.


Introduction and methods
Discussing the Old Irish glosses on Priscian's Latin grammatical textbook, Moran (2015, 136) made this thoughtprovoking statement "We may [...] question whether the Irish glosses are original compositions at all, or merely translations from inherited Latin sources."Previous studies on these matters, e.g., Lambert (1983), Killion (1992) or Bauer (2019d), have shown that there is no straightforward answer to this complex question and minute analyses of individual glosses need to be carried out.They have demonstrated that there are examples for translations from Latin to Old Irish but also the other way around.For the glosses found in St Gall, Stiftsbibliothek, MS 904 -a copy of Priscian's Ars Grammaticae -I have argued elsewhere that "not all of the Old Irish glosses […] are mere copies of original Latin glosses" (Bauer, 2019d, 17).
Is this also true for the Old Irish glosses on the Venerable Bede's De Temporum Ratione (= DTR)?The English monk's computistical opus magnum is commonly dated to 725 (cf.Wallis, 2004, xvi fn. 4).It is about measuring time and constructing a calendar.And computus is, after all, "nothing more than a complicated mathematical problem: how to find the date of Easter" (Wallis, 2004, xviii).Bede's work was widely distributed in the early medieval period; since these are complicated matters, it was accordingly heavily glossed.Within the bulk of transmitted manuscripts there are also important sources for early medieval Celtic languages (see the sources below) and in what follows, I will concentrate on those.To find answers to the question raised in the title, I am focussing on glosses in parallel transmission in the present study.Parallel glosses are annotations on the same lemma of the base text transmitted in different manuscripts.A first list of Celtic parallel glosses on DTR -concentrating mainly on the vernacular glosses of Angers 477 (Old Breton/Welsh) and the Karlsruhe Bede (Old Irish) -was offered by Lambert (1983, 121-127).He identifies the following levels of parallel transmission (Lambert, 1983, 120): -linguistic borrowings from Irish to Breton -glosses with the same contents over the same lemmas in the base text; in this case the parallelism sometimes goes as far as word-by-word translation/transposition -glosses with the same contents which appear in different locations of the base text.
Lambert concluded that a good portion of Angers' commentary (especially the glosses in hand B) originates from an Irish tradition. 2The present study goes one step further back in the Celtic glossing tradition on Bede's De Temporum Ratione.
It takes the Vienna Bede (= BVi.) as its main source, to research the genesis of the Old Irish glossing tradition on Bede's computistical work.This tradition has also influenced Angers 477's (vernacular) glosses to a large extent.Accordingly, I am focussing on parallel glosses with a vernacular Old Irish gloss in BVi.(the oldest of the four manuscripts of the corpus -see below) and glosses in another language (Latin or Old Breton/Welsh) in (at least one of) the other three manuscripts.
The study consists of two main parts: the section corpus and analysis presents and scrutinises the data.The parallel glosses as well as the lemmas they are glossing in the base text are analysed using the historical-comparative method from philology and historical-linguistics.The two final sections synthesise the results and hence discuss the genesis of the vernacular 2 "Pour nous, une bonne partie du commentaire d'Angers (notamment dans les gloses B) est empruntée à une tradition irlandaise" (Lambert, 1983, 128-29).
In my translation: 'In my opinion, a good part of Angers' commentary (notably the glosses [from hand] B) is borrowed from an Irish tradition.' 1 Part of this study was presented at the Tionól 2022 (1819 November, 2022; Dublin Institute for Advanced Studies).I thank the participants for helpful input in the Q&A session and furthermore Jacopo Bisagni, David Stifter and Sean Winslow for invaluable advice and suggestions.All disclaimers apply.

Amendments from Version 1
This revised version includes the extensive feedback received from the three very helpful peer reviews.Their recommendations helped to make the article more concise and added clarity to my arguments.The biggest changes are found in the section "Synthesis and results" in which I have also changed Figure 1, as well as the "Conclusions" to which I have added another table for more clarity.All in all the overall argument and conclusions have remained the same.
Any further responses from the reviewers can be found at the end of the article glosses in the Vienna Bede -the main research question of this paper.

Sources
The source manuscripts for this article are identical to those of the Gloss-ViBe corpus (Bauer, 2021(Bauer, -2023)), which is available under https://gams.uni-graz.at/query:glossvibe.allglosses.This corpus consists of all the glosses in the Vienna Bede and their parallels found in three other manuscripts.The manuscripts were chosen, because they share a high number of parallel glosses.It has long been argued that the Old Irish glossing tradition on Bede had a significant impact on the (vernacular) glosses in Angers 477.For the most part, the Latin glosses in the St Gall manuscript are very closely related to those of the Karlsruhe Bede (cf.Bauer, 2019c

Data collection
For the data used in the present paper I have narrowed down the Gloss-ViBe corpus to all those instances in which there is an Old Irish gloss in the Vienna Bede that has a parallel gloss in at least one different language -Latin or Old Breton/Welsh -in one of the other three manuscripts mentioned in the sources (Bauer, 2023).

Corpus and analysis
The presentation of the corpus is based on the layout I have established for Bauer (2019b).The parallel glosses are consecutively ordered as they appear in the Vienna Bede.They are arranged according to the DTR chapters they are found in.The base text is quoted from Corpus Christianorum Series Latina edition (= CCSL 123B) by Jones (1977) and its translation by Wallis (2004). 9See Bauer (2019c, 32-33) for a detailed discussion of the manuscript. 10]http://www.e-codices.unifr.ch/en/csg/0251/bindingA/0/Sequence-435(accessed 25 April 2023).
DTR 1: or timuil 'of the darkness' For an analysis of the Old Irish gloss of BVi.I refer to my discussion in Bauer (2017, 31-32), where I present the gloss as /dorchai/ nocte Lat.nocte 'night' only appears as an abbreviation n and is not to be read as part of the gloss itself.It is a correction of the main text, because the scribe skipped over it when he copied the text of DTR. ) means Lat.super 'on, upon, over'.Since annotating epogomena with 'superlunar' only makes sense when one knows of this made-up connection, it seems plausible that the Old Irish gloss in BVi. is a translation of the Latin found in the other two manuscripts rather than the other way around.
On one point, in fact, they differed from the Greeks, for while they intercalated after the end of the last month, the Romans intercalated, not at the end of February, but after its 23 rd day, that is, when the Terminalia was over.(Wallis, 2004, (1983) arguments it seems likely that the Old Breton gloss is a translation from the Old Irish ones.This example is furthermore interesting because the parallel glosses do not appear in the exact same position within the base text.In BVi. and Ang. the glosses appear over Lat.cendecusoctus 'one hundred and eighteen' hence providing a vernacular translation of the somewhat hard to read long Latin number.In contrast to this, the gloss in BCr.appears in the intercolumnar space with a signe de renvois linking it to the same number which is depicted with Roman numerals just two words afterwards (cf.also Jones' edition).It is also worth mentioning that cendecusoctus appears as cendecus octus at the line break in this manuscript.The gap caused by the continuation of the number in the next line makes the following interpretation possible.The glossator of the Karlsruhe Bede could have found the gloss in their exemplar and decided to put it into the intercolumnar space, because of the line break in the lemma it is glossing.The signe de renvois was only added later.Since the spelled-out number is divided by the line break the glossator (or somebody else) did not look at the original lemma anymore, but decided that the gloss is for the Roman numeral cxuiii which appears as the third "word" in the line.

Synthesis and results
Figure 1 shows the distribution of parallel glosses between the different manuscripts and records their languages.What immediately strikes the eye is that there are no Old Irish parallel glosses before chapter XVIIII.This is caused by the fact that there are only three Old Irish glosses on DTR in BCr.before chapter XVII. 19Parallel Old Breton glosses in Ang., on the other hand, exist throughout the chapters transmitted in the Vienna Bede (with a gap from chapter VIII to XIII -which does not mean that there are no Old Breton glosses within this range).It is furthermore interesting to see that the number of bilingual glosses is rather small: eight bilingual glosses stand against seventy monolingual glosses.But what does the presented corpus tell us about whether the vernacular glosses are originals or translations?A first step is actually to look at the bilingual glosses.Table 1 gives an overview of them.
We can see that most of the bilingual glosses (i.e.five) in our corpus appear in the Vienna Bede -Angers 477 has two instances and the Karlsruhe Bede only one.Taking a closer look at the data we see that in DTR 8, DTR 15, and DTR 25, the language switch happens before or after the Tironian note .i. which either stands for Lat.id est or OIr.ed ón.Bisagni (2013Bisagni ( -2014, 26) , 26) rightly states that such instances should not be counted as code-switches, because "there is no way of establishing with certainty whether the Irish and the Latin section were composed at the same time, and by the same person."This means that either the Irish or the Latin part might have been added at a later stage.Nonetheless, they are valuable cases for the present study, because at some point a glossator decided to add an Irish word or phrase to the otherwise Latin annotation.
For instance, it is interesting to note that while the Irish phrase in DTR 8 appears after the Latin part, DTR 15 and DTR 25 have the inverted order.And if we take a closer look at the latter two and compare them with the parallel glosses in Ang., we see that the latter manuscript has Latin words in the exact same position: Lat.festis vs. OIr.feli and Lat.egiptii vs. OIr.egiptacdae.Therefore, translation suggests itself.A direction for this is hard to determine -especially for DTR 15.In the case of DTR 25 we might tentatively suggest that Lat.egiptii in Ang.58a37h goes back to OIr. egiptacdae in BVi.4b44.58, because it does not appear in BCr.32b44.This suggestion finds support in DTR 4. Similar to DTR 25, BVi. and Ang. also share parallels which are not present in the other two manuscripts here.Recalling Lambert's (1983, 128-129) statement from the introduction, in which he mentions a strong Irish influence on the glosses of Angers (especially the vernacular ones), we can interpret Ang.50a7c amestidiou 'circular courses' as being a translation from OIr. fithissi 'circular courses'.This suggests that BVi.1a18.5 is an original composition.In analogy to this, it could also be tentatively argued that (the beginning of) Ang.58a37h in example DTR 25 is a translation from BVi. 4b44.58.Similar to the previous examples, the BVi.gloss in DTR 6 also starts with Old Irish and then continues in Latin.A direction of translation, however, is not possible to determine in this case.The same is true for the two bilingual glosses found in Angers 477.
Turning to the monolingual glosses now, I will concentrate on those glosses in which BVi. has an Old Irish gloss and at least one of the other manuscripts has a Latin one.The examples in DTR 9 and DTR 16 show a closer connection between BVi. and Ang., in contrast to the other two manuscripts.This confirms to what we can generally deduct from the data: until chapter XIII the glosses of BVi. and Ang.have a lot in common.However, once the Old Irish glosses appear in the Karlsruhe Bede, i.e. after chapter XVII, BVi. and BCr. are more closely connected.
Figure 1 visualises this very well, because we see that from chapter XVIIII onwards the two manuscripts always share parallel glosses in the same language and if one looks at the corpus presented above it shows that these are mostly verbatim.Further research is necessary to find possible reasons for the absence of Old Irish glosses in BCr.before folio 31 recto.
As already mentioned in the discussion of DTR 5 above, I show in a forthcoming publication that the Old Irish gloss is a translation from the Latin one.Since DTR 3 is quite similar in the sense that there are also two glosses only consisting of a verbal form, it seems plausible that BVi.1a16.4 is also a translation from the Latin gloss BCr.27a56.However, the peculiarities of the glosses mentioned above make any definite decision impossible.As already argued, a translation from Latin seems plausible for BVi.2b28.25 in DTR 14. Scholars like O' Sullivan (2004, 84), Moran (2015, 141) and Bauer (2019b, 52) have stressed the need for extensive editions of glossed manuscripts and the glosses in DTR 12 and 19 show how important they actually are.Only with the help of the parallel gloss in BCr. it was possible to find the correct reading of BVi.2a28.19a in the first example.In the second example the parallel gloss in Ang.helped to refine the grammatical analysis of BVi.4a2a.43 and BCr.32a12.The form of the substantive verb in these glosses (OIr.•mbí) has been wrongly interpreted as third person habitual present in previous scholarship.

Conclusions
There is no straightforward answer to the question of whether the vernacular Irish glosses are original compositions or translations from Latin glosses.As shown in Table 2, the discussion above features examples for both.There are two examples (DTR 5 and 14) for which a translation from Latin into Irish is almost inevitable.A translation in the same direction is also very likely for DTR 10.A rendering of originally Irish glosses in Latin is presumably the case in examples DTR 3 and 23.Especially the latter example shows that the strong Irish influence on the Old Breton/Welsh glosses found in Ang. as argued by Lambert (1983) could also be extended onto the Latin glosses in this manuscript.Future research will enable us to come to more informed conclusions on these matters.There are also five examples in the corpus which do not directly help to answer the research question, because they are Old Irish/Old Breton parallel glosses.Nonetheless, they demonstrate how closely connected the Celtic Bede manuscripts are.Until additional data is provided in the form of editions of more manuscripts, most of the examples stay ambiguous and therefore belong to the final column.
The fuzziness of the data is, however, not surprising.It always needs to be kept in mind that the early medieval gloss corpora are not homogeneous (cf.Bauer, 2019d, 17), but were layered and copied over a long time-span.The different linguistic and thereby also chronological strata of the Old Irish glosses on Priscian's Latin grammar Ars Grammaticae found in St Gall MS 904, for instance, have been the topic of several studies for more than a century now (cf.Lambert, 1996;Roost, 2013;Strachan, 1903) and yet no definite conclusions have been drawn to date.The issue of the genesis of the vernacular Celtic glosses is similarly complex.Since the Vienna Bede only survived the centuries as a fragment, the present article can only be seen as a pilot-study.Building on the presented methodology, more research should for instance be carried out on the Old Irish glosses in the Karlsruhe Bede and their potential Latin and Old Breton/Welsh parallels in Angers 477.This will bring further elucidation to the matters discussed here.Such studies, however, are only possible with extensive editions of early medieval manuscripts, including both vernacular and the Latin glosses.Only with such editions at hand will we be able to research the glosses as what they really are, a vital window into early medieval cultural and linguistic contact and the intellectual formation of Europe.7,8,9,11,13,15,16,17,19,21,22,24,25,26 Pádraic Moran Classics (School of Languages, Literatures and Cultures), University of Galway, Galway, County Galway, Ireland The article contributes to recent scholarship that investigates glosses on Latin manuscripts as evidence for early medieval multilingualism and language interaction.The corpus selected for study is clearly defined and well-organised.It expands previous work focusing on translation between Latin and Old Irish in the glosses by considering a third language, Old Breton.It also builds on an earlier study by Lambert (2005) by including an important third manuscript for Bede.
The data is clearly presented and the analysis will certainly be of interest to scholars of Old Irish and/or Old Breton.
The article needs improvement in the following respects: 1) The argument is somewhat inconclusive, both in the analysis of each group of glosses and in the overall conclusion.
For each group of glosses, the article should clearly state whether or not any information about direction of translation can be inferred.(This is left unclear e.g. for groups DTR 15,17,21,22,23,25,27.For DTR 17, even if the material is analysed in detail elsewhere, we need to know here what the conclusion is.) The overall conclusion should be more clearly presented.It begins with a very uncertain tone ('a bit of both'), but eventually turns to stronger evidence ('a translation from Latin to Irish is almost inevitable').It would be more useful to consider the strongest evidence first, then to consider its implications, before adding caveats.The final discussion of future possible research should be in a separate paragraph, for clarity.
2) The following are issues regarding interpretation: DTR 1: Gloss is surely intended for caeca, not nocte?
Open Research Europe DTR 2: 'although we do not see (it)'-why is 'it' in brackets, when there is clearly an infixed pronoun?
DTR 3: adrigiter is presumably a typo for ardrigiter?If a relative, is the translation 'by which they appear' necessary?Why not just 'which appear'?This could explained if the author fronted sidera in his (mental) translation: 'stars which appear to be shining lights'.
DTR 4: 'in Greek, circle is translated as apside' should be 'in Greek, absida is translated as circle' (for better clarify and mirroring the other translations following).'While Gr. αψίδα is usually only used to denote the architectural structure, the related form Gr. ἁψῑς can also mean 'wheel, hoop, disc'.'This is very confusing.They are presumably the same word ἁψίς (ἁψίδα is acc.sg.), which easily encompasses all of these meanings.There is no need for a separation here.The confusion really arises in Isidore (who says he is not sure what the nominative should be).'A connection with circle is therefore not too far-fetched.'Far from being far-fetched, it is the core meaning.The connection with 'bright' is the far-fetched one, since it is an invention by Isidore to service an etymology.
DTR 8: Latin nouns in oblique cases are supplied with prepositions elsewhere (e.g.DTR 9, 15, 16) so uno/una (die) should be here too ('with one day'?).'one Sunday' for una die dominico seems inadequate.Suggest 'with one day, a Sunday' (which more closely parallels the other glosses).Does we have to assume a progression from simple to more complex?What about the other possibility, that glosses were abbreviating their source material?Surely that's equally likely?
DTR 9: 'Since Old Irish does not have an ablative case…'.That is not relevant in this situation.The Latin lemma collectis 'after being collected' is indeed ablative, but the Irish gloss is not a substitution gloss.It instead provides supplementary information ('[collected] for an offering'), so does not need to mirror the case of the lemma.
DTR 20: dind leth ailiu (× 2) is idiomatic for 'on the other side'; see eDIL s.v.leth IV (c) (and cf.Latin ex parte).In which case, the distinction of prepositions discussed below seems not significant: they are both saying the same thing.
DTR 24: 'in contrast to'-suggest reword, since BCr. is just slightly more specific, not really contrastive.
3) Various minor errors and inconsistencies need to be corrected:

Chantal Kobel
School of Celtic Studies, Dublin Institute for Advanced Studies, Dublin, Ireland An interesting article that seeks to address the difficult question of the genesis of the vernacular glosses on Bede's De Temporum Ratione.Well worth publishing.This is a good case study of a small corpus of material which sets out the data clearly, followed by a detailed discussion of the results.
The following comments are intended to help strengthen certain points in the article, followed by some recommended minor stylistic changes.

Comments:
p. 4a: What is the relationship between the four copies of glosses?Perhaps a note following the description of the manuscripts would be helpful here.

Aaron Griffith
Utrecht University, Utrecht, Utrecht, The Netherlands The article examines early vernacular Celtic glosses and tries to answer the question as to whether they are original compositions or translations.This is a large and complicated question, so the article examines it within a limited corpus, that of the Vienna Bede.The question is also part of a larger discourse on glossing and glossing contexts, and this is indicated appropriately via citations of other relevant works and in the general approach.
The Vienna Bede is only a fragment, so there is a limited amount of data available to analyse.The conclusions are therefore quite modest.Nonetheless, there are some valuable observations and emendations suggested for individual glosses in the corpus (e.g.DTR 12 and 19).This article brings together such corrections with a well-informed overview of the situation surrounding glossing on DTR in a Celtic context.
The data presentation and analysis is brief, but generally clear, and the same goes for the synthesis.For the bilingual glosses under examination, there are a number of observations and suggestions for interpretation.These are all plausible, though not forcing (the author is clear on this, however, and is not over-interpreting the data).I do not think that stronger conclusions are possible or warranted at this time.The author's plea for further complete editions of texts is fullyjustified and a very reasonable end to the article.The article is a relatively short contribution to existing scholarship, but it presents results commensurate with that and is worthy of publication.
A number of specific comments follow (referring to page number and column letter): Page 3b bottom: "The present study goes one step further back in the Celtic glossing tradition on Bede's De Temporum Ratione.It takes the Vienna Bede (= BVi.) as its main source, to research the genesis of the Old Irish glossing tradition on Bede's computistical work which influenced the Angers 477's (vernacular) glosses to a large extent."There is a bit of ambiguity in this statement.Is the assumption that BVi is the primary source for Anger 477?Or is this to be proven?Page 4a top: for "chapter", rather "section"?
Page 4a: DTR3: Is BVi 1a16.4 mistyped?Should it not read ardrigiter 'by which they appear' (as in the main text)?Why is it translated as a prepositional relative instead of a regular relative 'which appear'?
Page 5a top: The explanation offered of Old Irish con•destis is presumably a concise summary of the argument in Bauer, forthcoming.Could this perhaps better be indicated directly?
Page 9a: DTR 17: though there is a reference to a full discussion elsewhere, it would be good to give some indication to the reader what one should take away from this gloss: direct connection but unclear direction?likely but unclear connection?
Page 12: Figure 1 presentation: The figure is effective in representing the data.I wonder if it would be more effective, however, if the manuscripts each got their own row in the data (with BVi on the bottom as now, then a row for BCr, then Ang, then Sg), rather than simply stacking them up as now.It would make it harder to see which individual DTR number had how many glosses (since there would now be gaps in the table), but since the numbers are low, this would not be a significant hindrance.The added benefit would be that it would be easier to see what individual mss.were doing with respect to the language of the parallel glosses under discussion.
Page 13a: When turning to the more numerous monolingual glosses, I wonder where some of the generalisations are coming from.That is, it is stated: "This reflects the general observation that the glosses of BVi. and Ang.have more in common until chapter XIII.However, once the Old Irish glosses appear in the Karlsruhe Bede, i.e. after chapter XVII, BVi. and BCr. are more closely connected."It is not clear to me where these observations are derived from.The data visualisation indeed is in agreement with these obsevations, but is this also the evidence or is it intended as confirmation of others' work?Reviewer Expertise: Early Irish, Middle Welsh, historical linguistics, language change I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Figure 1 .
Figure 1.Distribution of the parallel glosses.
Page 13b: To the references in the layers in the Sg.glosses, one might now add Griffith 2022 and Meng 2023.The statement in the text remains true, however.References 1. Griffith A: Examining linguistic layers in the St. Gall glosses, again.North American journal of Celtic studies.2022; 6 (1): 143-160 Publisher Full Text 2. Meng Qingyu: A Clustering Approach to Uncover Hidden Layers of Glosses in the St. Gall Priscian.Utrecht University BA thesis.2023.Is the work original in terms of material and argument?Yes Does it sufficiently engage with relevant methodologies and secondary literature on the topic?Yes Is the work clearly and cogently presented?Yes Is the argument persuasive and supported by evidence?Yes If any, are all the source data and materials underlying the results available?Yes Does the research article contribute to the cultural, historical, social understanding of the field?Yes Competing Interests: No competing interests were disclosed.

Badische Landesbibliothek, Augiensis pergamentum 167 (olim Codex Augiensis CLXVII) dates
Glosses on the base text appear in Latin and Old Irish.Digital images of the manuscript are available via the Badische Landesbibliothek. 84. St Gall, Stiftsbibliothek, MS 251 also dates to the first half of the ninth century. 9It contains Bede's De Natura Rerum and De Temporum Ratione in full, and the ending of De Temporibus -all annotated in Latin.High-resolution scans are available via the e-codices project. 10 ). 1. Vienna, Österreichische Nationalbibliothek, Codex 152985 (olim Suppl.2698) dates from the late eighth/early ninth century.The fragmentary manuscript -only four folios have survived -transmits (parts of) twelve chapters of Bede's De Temporum Ratione 3 .There are glosses in Latin and Old Irish.Images of the manuscript are available online from the Austrian National Library 4 and in the facsimile view of Gloss-ViBe 5 .2. Angers, Bibliothèque municipale 477 originates from either Brittany or North-East France 6 .A calculation found on folio 21a dates it to 897 (cf.Lambert, 2005).This composite manuscript contains, inter alia, Bede's De Temporibus, De Temporum Ratione, and De Natura Rerum.There are glosses in Latin, Old Breton/ Welsh, and "bretonised" Old Irish (cf.Lambert, 2005).A digital facsimile of the manuscript is online at the Bibliothèque Virtuelle des Manuscrits Médiévaux. 73. Karlsruhe, to the first half of the ninth century (cf.Bronner, 2013).Among other computistical works, it features De Temporibus, De Temporum Ratione, and De Natura Rerum.
Item De Eodem Si Quis Computare Non Didicit 'More on the same subject: for those who do not know how to calculate', XX.Quota Sit Luna In Kalendas Quasque 'What the age of the Moon is on any given first day of the month', XXI.Quae Sit Feria In Kalendas 'What day of the week it is on the Kalends', XXII.Argumentum De Qualibet Luna Vel Feria 'A formula for any Moon or weekday'.
While the Old Irish gloss features an adjective, the Old Breton parallel gloss in Ang. has a noun phrase consisting of preposition plus article and a noun meaning 'darkness'.The gloss in Ang.states the feriae are the days of the week, whereas the Old Irish gloss in the Vienna Bede stresses the fact that Pope Silvester instructed the clergy to have weekdays for 'the full space of a week' i.e. every single day.Therefore, a connection is possible, but not necessary.thirdkind of week occurs in the celebration of Pentecost; it is completed in 7 times 7 weeks, plus one, which is 50 days.[…]entertheHoly of Holies after the year's fruits of grain, wine and oil had first been col-As already mentioned for DTR 5 above, infinitives are a grammatical category alien to Old Irish.Therefore, the glossators had to find other ways of explaining them to Irish-speakers.In BVi.1d10.18, the Latin present passive infinitive lustrari (from Lat.lustrare 'to purify, circle, wander over, illuminate') is translated by the third person plural past subjunctive passive of the verb OIr.glanaid 'to cleanse, purify, purge'.For the Moon, which this year, for instance, is in its 17 th day on the nones of May [7 May] (Wallis, 2004, 43)Although DTR 12 is not an example of a vernacular/Latin parallel gloss, it is worth discussing here.Because so far BVi.2a28.19a has been read as Old Irish i•mbe 'in which you may be', a nasalising relative construction 15 featuring the preposition i 'in' and the second singular present subjunctive of the substantive verb.This would have somehow fitted the Latin gloss found in BCr., which has the preposition in, the relative pro-The same concept is rendered in Latin in BCr. and Sg. with the preposition Lat.super 'on, upon, over' plus the adjective lunaris 'lunar'.The substitution of Lat.super with OIr. for is commonly found in the glosses and also in the present corpus.In DTR 21, BVi.4b10.50 and BCr.32a57 have the Old Irish preposition with a Roman numeral for xi 'on the eleventh' where Ang.58a11b has super xi 'id.' (see below) 17 .While OIr. forescaidi is the only word of the gloss in BVi., the parallel gloss in the other two manuscripts offers an Isidorian etymology for the glossed lemma epogomena.This goes back to Gr. επαγομενος 'added on' and is used for the five extra days which are added to a year in the Egyptian calendar to make each year last 365 days.Only the etymology presented in BCr.
A lected in order[…]are commanded by the Law to be purified on the first, third and seventh day(Wallis, 2004, 35) DTR 8: BVi.1c38.16 uno .i. ond oenfiur 'one i.e. from one man' Ang.51a16e .i. una die dominico 'one Sunday' BCr.27d44 uno die 'one day' Sg. 57.37b i. uno die 'i.e. one day' The gloss in BVi.falls into two parts, a Latin one and an Old Irish one which are separated by the Tironian note for 'id est'.While the second part does not have a parallel in the other three manuscripts the first one -Lat.uno 'one' -occurs in all three of them.It explains the meaning of monade found in the base text.Lat.monas goes back to Gr. μονας 'unit'.The gloss in BVi.simply notes that 'one' is meant with monade.BCr. and Sg.add the word for day and the longest gloss, i.e.Ang.54a16e, states that the 50th day (mentioned in the base text) is a Sunday.tion OIr.do 'to, for' plus the dative singular of OIr.idbart '(act of) offering, sacrifice', the verbal noun of OIr.ad•opair 'offers, sacrifices'.Ang.51a22b features the ablative singular of Lat.sacrificium 'something made sacred or given to a deity, sacrifice'.It has to remain unclear whether or not the glosses form translations of each other.The other three manuscripts all have mundari, also a present passive infinitive (of mundare 'to clean, cleanse').BCr.offers two more passive infinitives with synonyms meaning 'to be consecrated, dedicated, cleansed, washed'.In this example the grammatical analysis of BVi.1d10.18 allows both, a translation of the lemma in the main text, as well as a translation of the Latin glosses.A translation in the opposite direction is equally possible, though less probable.VIIII.De Hebdomadibus Septuaginta Propheticis CCSL 123B, 305 … embolismos uero menses, qui de annuis xi epactarum diebus adcrescere solent, non lege patria tertio uel altero anno singulos adiciens DTR 11 … […] did not include in the second or third years (as tradition decrees) the embolismic months which normally accumulate from the eleven days of the epact of every year.(Wallis, 2004, 36) DTR 11: BVi.1d26b.18b.i. indeud ogdato ocus circuil 'i.e. in the end of the octad and the [entire] cycle' XI.De Mensibus CCSL 132B, 315 Quia uidelicet luna, quae praesenti DTR12 anno, uerbi gratia, per nonas Maias septima decima existit noun quo and also a form of the substantive verb.The latter however is the third singular perfect indicative 'he/she/it was', which is in concordance with the nominative singular Noe 'Noah'.The last word of BCr.29b4 led me to the new (and most likely correct) reading of the parallel gloss.There is a crease in the manuscript at the beginning of BVi.2a28.19a which makes the beginning of the gloss a bit uncertain, but it is very likely the Tironian note .i. 'i.e.'.The rest of the gloss noe is, however, clearly readable.Hence, this example shows the importance of researching glossing traditions as a whole, because only with the help of parallel glosses can certain glosses be understood and correctly interpreted.The base text of chapter XI talks about the age of the moon on the nones of May and gives Noah's ark as an example, because that's the date that Noah went into the Ark.Bede goes on to say that the moon is in its 17 th day on the nones of May [May 7] and it will be 27 days old on the day before the nones of May in the following year.The two glosses tell the reader that Bede is talking about the year in which Noah went into the Ark.This stands in opposition to the interpretation of modern scholarship.It is the common opinion nowadays that Bede is actually talking about his present. 16Indeed, the passage is one of the examples that is used to date the composition of DTR, because the moon would have been 17 days old on May 7 in the year 722. 15See DTR 19 below for a similar construction. 16Cf.Wallis (2004, 43 fn.121). of nature has taught, the Moon plainly completes the zodiac in 27 days and 8 hours, but its proper course is 29 days and 12 hours, setting aside the calculation of the "leap of the Moon".(Wallis, 2004, 43-44) DTR 13: BVi.2a35.22 .i. reim ṅgrein 'the course of the sun' Ang.53b11a ad solem sequendum tamen 'however, to follow the sun' Although the two glosses bear very similar semantics and appear at the same position within Bede's textbook, the exact connection is not clear.CCSL 123B, 318 Quem decimo kl.Septembrium die terminantes, residuos quinque dies epagomenas DTR 14 uel intercalares siue additos uocant… this [last month] ends on the 10 th kalends of September [23 August], and they call the remaining five days epagomena -"intercalated" or "added" (Wallis, 2004, 45) DTR 14: e. superlunar because "mene" means "luna" and "epo" "super"' The gloss in BVi.only consists of the nominative plural forescaidi 'superlunar'.This hapax legomenon is a compound of the preposition OIr. for 'on, upon, over' and the adjective OIr.éscaide 'lunar' -an adjectival formation to OIr. éscae 'moon, month'.29c1 and Sg.63.21 helps to understand the Old Irish gloss in the Vienna Bede.It states that epogomena means 'superlunar' because mene (= Gr. ) means Lat.luna 'moon' and epo (= Gr. fiebant guilou termini 'before in the days of intercalation the feasts of Terminus took place'.Lambert (1983, 122)only records the final two words of this gloss and accordingly connects the two Celtic glosses although they do not appear on the same lemma of the base text.The ways in which the four glosses of DTR 15 are connected remain uncertain at this point.In olden times, the responsibility for observing the first appearance of the new Moon and of announcing its sighting to the royal sacrificing-priest was delegated to a minor priest.(Wallis,2004,50-51)Stifterhasrecentlydiscussed the interconnections of OIr.félire and OBret.guiler,guilerou,guileri and OW gueleri all bearing the same semantics 'calendar'.He concluded that the latter are "learned, erudite adaptations […] of OIr.félire"(Stifter, 2022, 4).DTR 18 is, therefore, not necessarily an example of direct translation/transposition from Old Irish to Old Breton, but reflects common practice.(2021)translate/analysetheform of the substantive verb OIr.•mbi in BVi. and BCr. as third person singular habitual present.In the light of the parallel gloss in Ang. this breakdown should be reconsidered.In analogy to Lat. es 'you are' in Ang.57b26d, the verbal form in the Old Irish glosses should also be interpreted as second person singular habitual present.All three glosses therefore directly address the reader.While a connection between the Latin and the Old Irish glosses is very likely, the direction of translation is, unfortunately, uncertain.The constructions in the Irish as well as the Breton glosses are very similar.The only real difference is the noun: OIr.leth vs. OBret.parth.This might be caused by the fact that the native cognate of OIr.leth, OBret.let is only attested in compounds(cf.Fleuriot, 1964, 241).Since there is a strong Irish influence on the vernacular glosses of Ang., it seems likely that Ang.57b34a is translated from the Old Irish glosses.Si uis scire quota est luna in kl.iunias anno tertio, tene regulares xii, adde epactas DTR 21 anni illius xxii, fiunt xxxiiii.Tolle xxx, remanent iiii.Quarta est luna in kl.memoratas.Quod si quis obiecerit DTR 22 uel huius uel praecedentis argumenti alicubi ordinem ordinem uacillare… If you want to know what Moon it is on the kalends of June in the third year, take the regular 12, add the epact for that year -22 -and that makes 34.Subtract 30, and 4 remain; on the kalends in question, the Moon is four days old.Should anyone object that the order of either this or the preceding formula is shaky at any point […] (Wallis, 2004, 66)at that point, it is necessary that the Moon of July in that year have twenty-nine days and never more, one day having been removed because of the "leap of the Moon".(Wallis,2004,67)Although the gloss is very hard to decipher in BVi., it seems that it has the Tironian note .i. at this position, because at least the first full stop and the i can be read.This speaks in favour of the Tironian note .i. and against an interpretation as 'in'.The direction of translation for the adjectives introducing the glosses of BVi. and Ang.have to remain unclear.In contrast to the two Old Irish glosses which have the numeral particle a, the gloss in Ang.features the third person singular present indicative of the copula in this position.Otherwise the glosses are very similar and following Lambert's 'i.e. of the terminal feast'Although not using the exact same words, all glosses feature the phrase 'feasts of Terminus'.In contrast to BVi., Ang. and Sg., BCr. has a more elaborate gloss which also features hic est plutonis.It is also noteworthy that in Ang. a part of a marginal gloss a few lines above (Ang.54abis19f)featuresanOldBreton and Latin version of the Old Irish gloss found in BVi.: ante in dies interkalationis As I have stated elsewhere(Bauer, 2019c, 41)the glosses in BVi. and Ang. are very likely connected and might even be translations of each other.A direction for this translation, however, cannot be established.CCSL 123B, 344… aperto codice DTR 18 nota literam quae eidem sit praeposita diei DTR 19 … […] open the [calendar-]codex, and note the letter prefixed to that day.(Wallis, 2004, 63) The glosses annotate Lat.diei 'day' in what reads in Wallis' (2004, 63) translation: "[s]o if you wish to know what sign or what part of the month the Moon is in on any given day of the year, open the [calendar-] codex, and note the letter prefixed to that day."Both Stokes and Strachan (1901-1903) and Stifter et al.CCSL 123B, 348 Si enim ipsum argumentum iuxta Aegyptios a Septembrio mense ubi principium est anni eorum inchoaueris, necesse est ut luna Iulii mensis eo anno xxviiii dies ut numquam alias habeat, uno uidelicet ratione saltus amisso DTR 25 But if you start [to use] this formula at the month of September, after the manner of the Egyptians, whose year begins XXII.Argumentum De Qualibet Luna Vel Feria CCSL 123B, 352 adde xiiii fiunt cxxviiii; partire per lviiii (quinquagies nouies bini cendecusoctus DTR 27 ), tolle cxviii, remanent xxviii.

Is the work original in terms of material and argument? Yes Does it sufficiently engage with relevant methodologies and secondary literature on the topic? Yes Is the work clearly and cogently presented? Partly Is the argument persuasive and supported by evidence? Partly If any, are all the source data and materials underlying the results available? Yes Does the research article contribute to the cultural, historical, social understanding of the field? Yes Competing Interests:
No competing interests were disclosed.

have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
https://doi.org/10.21956/openreseurope.17283.r33549© 2023 Kobel C.This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Some minor stylistic points:
Cf. the comment on p. 13a 'the gloss in BVi. is a copy from another exemplar …'.Does this imply that other shared glosses between BVi./ Ang..5a'BVi.1a16.4adrigiter'>recte ardrigiter.p. 7b7.8: add reference to Thesaurus Palaeohibernicus here for the former reading of i mbe.p. 7b20 'but it is very likely the Tironian note .i.'.Yes, I would agree.In fact, I can make out the letter i and following punctum, with only the first punctum obscured by the crease in the manuscript.p. 8a1 'BVi.2a35.22.i. reim nġrein'.It would be worthwhile having a note here on the form nġrein, which is translated as a gen.sg.! Is this the correct reading?Thes.Pal.has ngreine..11a1'it seems that it has the Tironian note .i. at this position, because at least the first full stop of it can be read'.Is that really the case?If only the first stop can be read (and without following i.), what's not to say that this is a full stop?I find most of this gloss is almost entirely illegible apart from the final abbreviation for kalendae.This should be made clear here in the first instance, rather than on p. 13.Avoid the use of the passive, e.g.p. 4 'The two final chapters are synthesising the results' > 'the two final chapters synthesise the results'; p. 5a24 'I am showing' > 'I show'.Some sentences are quite long, extending over four lines of text, see for instance p. 11 col.b beg.'The signe de renvois was only added …' (this sentence has over 50 words!).Rewrite as shorter sentences for ease of reading.
p. 5a 'BVi.1a9b1 dorchai'.Should the complete gloss (dorchai nocte) be given here to fully express the adjectival force in this instance and that it is not misconstrued as a substantivised adjective?This is what is printed in Ériu 67 (2017, 31) after all.ppp.12b14 'Which suggests that BVi.' > suggestion 'This suggests that …'.p. 12b20 'where BVi. has the Tironian note for id est'.But this reading is unclear in the manuscript (which you do mention on the following page).Perhaps one should be more circumspect here.Also, could the exemplar have had i + n-stroke for Lat.in, like in Ang.? The n-stroke may have been lost in transmission.

Is the work original in terms of material and argument? Yes Does it sufficiently engage with relevant methodologies and secondary literature on the topic? Yes Is the work clearly and cogently presented? Yes Is the argument persuasive and supported by evidence? Yes If any, are all the source data and materials underlying the results available? Yes Does the research article contribute to the cultural, historical, social understanding of the field? Yes Competing Interests:
No competing interests were disclosed.

have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
https://doi.org/10.21956/openreseurope.17283.r33551© 2023 Griffith A. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.