The Trouble with Prefixes

Consummatum est, inclinato capite tradidit spiritum.
‘”It is accomplished”, he bowed his head and gave up his spirit.’

(Homiliae XL in euangelia, homily 37.9, Gregory the Great)

These were the last words of bishop Cassius of Narnia as recorded in Bodleian Library, MS. Laud Misc. 429. The manuscript contains Homiliae XL in euangelia by Gregory the Great (see Bodleian online). Cassius of course quoted from the Bible, Jesus’ last words according to John 19.30. The quotation appears at the end of folio 149v which also contains two scratched glosses in the upper margin, barely visible to the naked eye: ‘[…] braht’ and ‘upbraht’. At first glance, the glosses and the quote have no obvious connection. They stand on opposite sides of the folio (fig. 1) and, semantically, the two similar glosses do not seem to relate to any Latin on the page. It was only when the glosses were captured in a detailed image with the Selene scanner of the ARCHiOx project (ARCHiOx: research and development in imaging) that it was possible to learn more about their origin and meaning.

Fig. 1: Folio 149v of MS. Laud. Misc. 429 (Oxford, Bodleian Library). The position of the scratched glosses is marked in red, the quote by Cassius of Narnia is marked in green.

With the Selene scanner, a clear image of the 3D surface of the object can be created. This is particularly useful for scratched glosses, translations and comments not written with a pen but impressed into the parchment with a stylus. Scratched glosses can usually only be seen by shining a torch onto the parchment at a very shallow angle. But with the new recording system the glosses can be made visible within the context of ink glosses and the main text. Corrections in the main text also become much clearer. Suddenly a sequence of short horizontal scratches become visible, showing where the scribe erased ink with a small knife (fig. 2), apparently to adjust word endings and punctuation. Furthermore, there are longer, finer lines crossing the page that stem from preparing the parchment.

Fig. 2: 3D‑render of the scratches on folio 149v recorded with the Selene scanner. In the top margin, the lexical scratched glosses are visible. On the bottom of the picture are scratches which stem from text erasures.

In the recording, the lexical scratched glosses on folio 149v are clearly visible, except for the first letters. They read (1) ‘[…] braht’ and (2) ‘upbraht’. The words appear right next to each other but are divided by a clear gap. These are not the only lexical glosses in the manuscript. Throughout the text, there are Latin and Old High German glosses, for example the Old High German word ‘agaleizor’ in the right margin of folio 159r and the Latin ‘lapidem’, a few lines below and interlinear (see the digitised manuscript and Hofmann 1963, p. 144). As for the glosses on folio 149v, the second half of the words gives clues as to their language. ‘‑braht’ is the past participle of the Old High German verb ‘bringan’ (see AWB 1,1384), with the basic meaning ‘to bring’ (or in case of the participle, ‘brought’). The first part of the words should therefore be a prepositional prefix which modifies the basic verb. The first part of gloss (2), ‘up’, is known as a prefix – but not in Old High German. The Old High German equivalent of the word is ‘ûf’, with the ‘‑p’ undergoing the Second Consonant Shift (up > uf). That the gloss still has the ‘‑p’ shows that it must belong to a language which did not go through this sound change. To explain this mix of linguistic features, it is necessary to consider the history of the manuscript.

MS. Laud Misc. 429 was written in a German writing centre, possibly Fulda, at the beginning of the ninth century (Bischoff & Hofmann 1952, p. 58; Mairhofer 2014, p. 680). It was in Würzburg from the fifteenth century at the latest and was given to the Bodleian in 1637 (Mairhofer 2014, p. 680). However, it is not unlikely that the manuscript came to Würzburg much earlier because of the intertwined history of the two monasteries. In the eighth century, Saint Boniface and his missionaries came from England and founded several monasteries, including the Würzburg cathedral chapter and Fulda as part of its diocese (Bischoff & Hofmann 1952, p. 5). As they were closely related, the monasteries frequently exchanged manuscripts in the eighth and ninth centuries, or would order manuscripts from each other’s scriptoria (see Bischoff & Hofmann 1952, pp. 142, 168). Fulda was therefore subject to the same Anglo-Saxon influence that has been widely researched for the Würzburg monastery. Anglo-Saxon traces can be seen palaeographically and linguistically in the manuscripts created during that time (Bischoff & Hofmann 1952, pp. 5ff.; Hofmann 1963, pp. 33–4). Written long after Saint Boniface’s death, when the main impact of the Anglo-Saxon mission had already waned, MS. Laud Misc. 429 still shows signs of Anglo-Saxon script (‘Symptome dafür sind die leicht spachtelförmigen Oberlängen, die mit dreieckigem Ansatz beginnenden, tiefgespaltenen r der rcc-Ligatur.’, Bischoff & Hofmann 1952, p. 58). Ties between early medieval scriptoria of the Würzburg diocese and Anglo-Saxon writing tradition remained strong over the centuries and MS. Laud Misc. 429 is evidence for that.

Fig. 3: Close‑up of gloss (2) ‘upbraht’.

On a different note, there is another, though rather minor possibility for the origin of the scratched glosses. Apparently, there have been exchanges between the Low German and East Franconian region since the Old Saxon Heliand was written under consideration of texts which stem originally from Fulda (Schubert 2013, p. 213). Linguistically, the scratched glosses are acceptable Old Saxon forms. Old Saxon did not undergo the second consonant shift, hence the ‘up‑’, and ‘braht’ is the past participle of ‘brengian’, the Old Saxon equivalent of Old High German ‘bringan’ (Gallée 1993, §408). Even though there is no evidence of an Old Saxon verb ‘upp‑brengian’, ‘upp‑’ is known as verbal prefix (Tiefenbach 2010, p. 431). It is similar with the possible prefixes of gloss (1) which will be considered later. Arguably, these could be two Old Saxon hapax legomena (words which are only evidenced once), however, connections between the scriptorium of Fulda and Old Saxon scribes are hardly documented. On the other hand, the link to Anglo‑Saxon writing is not only supported by other manuscripts from Fulda and by the palaeographical characteristics of the script of MS. Laud Misc. 429 but also matches the linguistic evidence seen in gloss (2): ‘up’ (fig. 3) can very well be an Old English form (predecessor of modern English ‘up’) and the combination of an Old English prefix with an Old High German verb stem is a typical result of Anglo-Saxon influence in German writing centres. Since none of the ink glosses show Old English features (see Hofmann 1963, pp. 114–5) it could be that ink and scratched glosses stem from different scribes who followed different writing traditions.

Fig. 4: Close‑up of gloss (1) ‘[…] braht’. The prefix could be read as ‘hu’, but note the difference between the first letter and the penultimate letter ‘h’.

The prefix of gloss (1) is not as easy to decipher. The first letter could be an ‘h’, even though it looks quite different from the ‘h’ in the second half of the word (fig. 4). The latter has a very round curve, while the former shows a sharp bend. And neither in Old English nor Old High German is ‘hu’ a documented verbal prefix.

Fig. 5: Prefix of gloss (1) ‘[…] braht’ with the possible reading ‘zu’ in comparison with the ‘z’ in an ink gloss on fol. 159r (image was rotated to align the letters).

Another possible reading is palaeographically less clear but would make sense lexically: ‘zu’. The lines of the first letter could be a crooked ‘z’ (compare the ‘z’ of ‘agaleizor’ on folio 159r), which, to be fair, would miss some strokes (fig. 5). But an Old High German prefix ‘zu‑’ (or ‘zuo‑’) is indeed recorded in combination with ‘bringan’ (AWB 1,1405; the Old English equivalent would be ‘tó‑’, see Bosworth‑Toller online). Both readings assume that the second part of the prefix is a ‘u’. However, the curves of the letters in ‘braht‑’ are very round and those of the potential ‘u’ are not. Because scratched glosses had to be scratched into the parchment with a stylus which was not a very reliable writing instrument, letter shapes could be distorted (Glaser & Nievergelt 2009, p. 207). Differences between the letters within a gloss could therefore happen and could have a number of reasons. The gap after the prefix, for example, could mean that the scribe paused and then held the stylus differently. But it is curious that the letters of the second part are all very neatly rounded and the first ones are not.

Fig. 6: Prefix of gloss (1) ‘[…] braht’ The image on the right shows the same prefix with mark‑up that illustrates the possible reading ‘vul’, excluding the strokes that connect the letters.

A reading which would take the sharp bends of the beginning a bit more into account would be ‘vul’. Just as in gloss (2), this prefix would be Old English, the Old High German equivalent being ‘fol’. Old English ‘full’ is recorded as a verbal prefix (e.g. ‘fullbétan’, Bosworth-Toller online) and the Old High German variant even exists in combination with ‘bringan’ (‘fol(la)bringan’, AWB 3,1049). However, this reading would assume a ‘v’ with an ascender and an ‘l’ which is missing one (fig. 6). Just as ‘zu braht’, it is palaeographically odd but seems lexically sensible.

The meaning of gloss (1) can be narrowed down by semantically interpreting gloss (2). Both glosses end in ‘braht’ which suggests that they relate to each other. Double glosses are often either synonymous and were meant to provide the reader with lexical variants, or they give alternative semantic interpretations of the lemma. This could be achieved by altering the prefix of a verb. Old English ‘up’ and ‘bringan’ together can have the meaning ‘bring it to pass’, in which ‘up’ is ‘marking effectual action’ (see Bosworth-Toller online; here as an adverb related to a verb, but also recorded as a prefix). In Old High German, only one instance is known in which ‘uf’ appears together with ‘bringan’, in the Muspilli: ‘die pringent sia sar uf in himilo rihi’ (Steinmeyer 1916, p. 66; see AWB 1,1391), and here it means ‘to take (someone) up (to somewhere)’. The form ‘upbraht’ is a past participle. Assuming that the gloss copies the grammatical form of its Latin lemma to translate it in context, it could correspond to Latin consummatum from the aforementioned bible quote. The Latin would fit the possible meaning ‘bring it to pass’, or rather ‘brought to pass’. Their position in relation to each other is rather unusual as they are at opposite ends of the page – glosses are mostly in direct proximity to the word they translate. However, the word summarises the whole tale of Cassius who awaited his death for years after hearing a vision from one of his priests. His last word, ‘accomplished’, could relate not only to his death but also to the long period of waiting. Glossing it on top of the page is similar to a headline.

In this regard, gloss (1) can be expected to have a similar meaning. Of the three presented readings, ‘hu braht’ is the least fitting one. The adverb exists only in Old English and is the predecessor of modern English ‘how’. It references the quality of a verb (Bosworth-Toller online) which does not fit very well in this context. The second variant, ‘zu braht’, works better. The Old High German verb ‘zuobringan’ can have the meaning ‘to bring about’ (AWB 1,1405) – even though the according Old English prefix does not match (‘a prefix denoting separation, division’, Bosworth-Toller online). And finally, the last reading ‘vul braht’ is semantically closest to ‘upbraht’. Old High German ‘fol(la)bringan’ means ‘to finish something, to accomplish something’ (AWB 3,1049). Old English ‘ful’ is a verbal prefix which ‘denotes the fulness, completeness or perfection of the meaning of the word with which it is joined’ (Bosworth-Toller online). Judging from these interpretations, the two glosses could have been designed to give lexical variants which are broadly synonymous for the Latin lemma.

Fig. 7: Profiles of the scratches of gloss (1) on the left and gloss (2) on the right as recorded with the Selene scanner. The line in the glosses marks where the profile was measured.

Again, the ARCHiOx recording reveals more information about the motivation behind the double glosses. Measuring the depth and width of the scratches of the two glosses in the 3D-image shows that their profiles do not match (fig. 7). That means that it is very likely that they were not written in one go. Either the scribe, the writing instrument or the date of writing changed, or possibly all three at once. Whatever the reason for the changing profile was, there was most definitely an interruption between writing the two glosses. Maybe one of the glosses seemed unsufficient to a glossator to translate the lemma, maybe someone working with the text wanted to give a variant of the translation – or maybe a later reader had the same problems in identifying the first gloss as I had and decided to add a more legible translation. With the help of the ARCHiOx recordings, it is possible to gain much more information about a fascinating linguistic phenomenon. The detailed images can paint a clearer picture of how people in the Middle Ages worked with texts. In the case of MS. Laud Misc. 429, the glosses can not only be linked to a rich history of language exchange, but we now have proof that that the manuscript was the subject of work processes that are much closer to today’s way of studying than one would think.

Image credit

Fig. 1: Bodleian Library (Oxford, Bodleian Library MS. Laud Misc. 429,, accessed 1 September 2023), with mark‑up by the author

Figs. 2–4: John Barrett, ARCHiOx

Fig. 5: left: John Barrett, ARCHiOx; right: Bodleian Library (Oxford, Bodleian Library MS. Laud Misc. 429,, accessed 1 September 2023), rotated detail by the author

Fig. 6: John Barrett, ARCHiOx, with mark‑up by the author on the right image

Fig. 7: John Barrett, ARCHiOx


AWB = Althochdeutsches Wörterbuch. Auf Grund der von Elias v. Steinmeyer hinterlassenen Sammlungen im Auftrag der Sächsischen Akademie der Wissenschaften zu Leipzig. Bearbeitet und herausgegeben von Elisabeth Karg-Gasterstädt und Theodor Frings. Leipzig 1952-2015ff., (accessed 1 September 2023).

Bischoff, B. & Hofmann, J. (1952): Libri Sancti Kyliani. Die Würzburger Schreibschule und die Dombibliothek im VIII. und IX. Jahrhundert. Würzburg: Ferdinand Schöningh.

Bodleian online = A catalogue of Western manuscripts at the Bodleian Libraries and selected Oxford colleges, (accessed 1 September 2023).

Bosworth-Toller online = Joseph Bosworth, An Anglo-Saxon Dictionary Online, edited by Thomas Northcote Toller, Christ Sean, and Ondřej Tichy. Prague: Faculty of Arts, Charles University, 2014. (accessed 1 September 2023)

Gallée, J.H. (1993): Altsächsische Grammatik. 3rd ed. Tübingen: Niemeyer.

Glaser, E. & Nievergelt, A. (2009): ‘Griffelglossen’, in Bergmann, R. & Stricker, S.: Die althochdeutsche und altsächsische Glossographie. Ein Handbuch. Vol. 1. Berlin/New York: De Gruyter, 202–229.

Haubrichs, W. (2013): ‘Volkssprachige (theodiske) Schriftlichkeit in Fulda (8.–11. Jh.)’, in Schubert, M.: Schreiborte des deutschen Mittelalters. Skriptorien – Werke – Mäzene. Berlin/Boston: De Gruyter, 196–215.

Hofmann, J. (1963): ‘Altenglische und althochdeutsche Glossen aus Würzburg und dem weiteren angelsächsischen Missionsgebiet’, Beiträge zur Geschichte der deutschen Sprache und Literatur (Halle) 85, 27–131.

Mairhofer, D. (2014): Medieval Manuscripts from Würzburg in the Bodleian Library, Oxford: A Descriptive Catalogue. Oxford: Bodleian Library.

Steinmeyer, E. (1916): Die kleineren althochdeutschen Sprachdenkmäler. Berlin: Weidmannsche Buchhandlung.

Tiefenbach, H. (2010): Altsächsisches Handwörterbuch. Berlin/New York: De Gruyter.

One Reply to “The Trouble with Prefixes”

Leave a Reply

Your email address will not be published. Required fields are marked *