An Armour of Sound. ‘Sancte Sator’ (‘Carmen ad Deum’) and its German Gloss

Abstract

The paper addresses the Latin poem Sancte sator / Carmen ad deum, which belongs to the early medieval tradition of loricae, and an Old High German gloss accompanying it in the manuscript clm 19410 (Munich BSB) from the 9^th century. The focus of the article lies on different dimensions of materiality playing a crucial role in the lorica genre and in Sancte sator in particular; that the German gloss refers and reacts to these material aspects is an important finding. The basic physical side of language (sound), the physical meaning of the literal content (source domain of thing metaphors), and the physical interaction of gloss and poem (with the options of the gloss leaving the poem untouched, joining it or ‘cutting’ it) are essential features. The last aspect will allude to the broader question of whether and how medieval glosses treat the glossed texts as material objects.

Abstract (German)

Das lateinische Reimgebet Sancte sator / Carmen ad deum steht in der frühmittelalterlichen Tradition der loricae; in der Handschrift clm 19410 (BSB München) aus dem 9. Jahrhundert wird es von einer althochdeutschen Glosse begleitet. Der Aufsatz fragt nach unterschiedlichen Dimensionen von Materialität, die für die Gattung der loricae und im Besonderen für Sancte sator zentral sind; es lässt sich zeigen, dass die Glosse auf diese physischen Aspekte Bezug nimmt und reagiert. Wesentlich sind die grundlegende physische Dimension der Sprache (Klang), die materielle Bedeutung der Signifikanten (Herkunftsbereich der Dingmetaphorik) sowie die materielle Interaktion von Glosse und Gedicht (welches von der Glosse unberührt bleiben, begleitet oder ‚zerschnitten‘ werden kann). Letzteres führt zu der Frage, ob und inwiefern mittelalterliche Glossen die Auctortexte als materielle Objekte behandeln.

Table of contents

The poem Sancte sator, named after its incipit, but also known as Carmen ad deum for its title in the first modern edition,¹ dates back to the 7^th or 8^th century. It draws on an originally Irish tradition of poems called loricae (sg. lorica). In the manuscript clm 19410 (Munich BSB) from the 9^th century, Sancte sator is accompanied by an Old High German gloss. In the following, I would like to show (1) how materiality is crucial to the lorica-genre and, in this particular case, to Sancte sator; (2) how the special sound of Sancte sator can be understood as another dimension of materiality; and (3) how the Old High German gloss complements or reacts to these aspects. The latter will allude to the broader question (4) of whether and how medieval glosses treat the “text proper”² as a material object.

Prayers as Breastplates: Loricae

Loricae are Christian apotropaic prayers of an originally insular tradition in Latin and Celtic languages (a famous example is the Irish Lorica of St. Patrick); they also influenced medieval Anglo-Saxon poetry.³ Lorica means ‘breastplate’, and often this very word appears in the texts, among other lexemes of armour. The prayers usually begin with an invocation of the Trinity,⁴ and then they call up pieces of armour in order to save the speaker’s body and soul from evil. Sometimes a litany of every body part, even hair and finger nails, is placed under the protection of ‘God’s shield’ (e.g. in the Lorica of Laidcenn). The conceptually central term lorica was “extended […] to refer to the prayer and the act of its recitation […] by scholars and the contemporary poets”⁵ – so the poem itself is a lorica. A fitting German term is “Panzer-Lied”.⁶

The main source of this idea is the Bible, namely verses containing figurative armour such as Eph. 6,11–17 (mentioning arma Dei ‘God’s armour’; loricam iustitiae ‘breastplate of righteousness’; scutum fidei ‘shield of faith’; galea salutis ‘helmet of hail’; gladium Spiritus quod est verbum dei ‘sword of the Spirit which is God’s word’) or 1Thess. 5,8 (loricam fidei et caritas ‘breastplate of faith and love’; galeam spem salutis ‘as a helmet the hope for salvation’).⁷ Inspired by these biblical images, the protective function of a lorica poem is achieved by the invocation of protective armour. However, in this very respect, the medieval texts differ considerably from the biblical passages.

The biblical texts make clear that armour vocabulary is only used metaphorically because the terms are disclosed by genitive phrases (or something similar) which explain their actual meaning. For instance, in the phrase loricam fidei et caritas (‘breastplate of faith and love’),  lorica is unmistakably a metaphor just characterizing fidei et caritas which the text is actually directed at. In medieval loricae, such disclosures are often remarkably reduced or even omitted, e.g.:

Deus […]
undique me defende potentia
Meae gibrae pernas omnes libera
tuta [B: tua] pelta protegente singula
Ut non tetri demones in latera
mea librent ut soleant iacula
Gigram cephale cum iaris et conas
pattham liganam sennas atque michinas
Cladum crassum madianum talios
bathma exugiam atque binas idumas

(‘God, […] from all sides defend me with power, deliver the whole trunk of my body with thine own protecting shield that foul demons may not hurl, as is their wont, their darts at my flanks, skull, head with hair and eyes, forehead, tongue, teeth and nose, neck, breast, side and reins, thighs, bladder and two hands.’)

(Lorica of Laidcenn/Gildas, paragraphs 15–19)⁸

The case this article focuses on, Sancte sator, at first uses weapon terms metaphorically (v. 12: et piacla   dira iacla [‘and sins, the cruel darts’]; v. 15: carnis nexu [‘bonds of flesh’])⁹, but, when introducing the protective armour, it soon spares metaphor clarifications, e.g.: Christi umbo   meo lumbo / sit, ut atro   cedat latro (v. 16f.) (‘May Christ’s shield be on my loins so that the robber with his dark may yield’); pater parma   procul arma / arce hostis (v. 19f.) (‘father, with the shield ward off the enemy’s weapons far away’). As in the Lorica of Laidcenn (see above), these armour terms are not presented as rhetorical devices. Whereas the biblical texts describe the invisible values as armour, the medieval loricae, as apotropaic prayers, implore the armour to be there, or we could say: implore it to materialise in order to achieve the protective function of the poem. As researchers have often pointed out, “the loricae are as much magical as they are religious”¹⁰; to Haubrichs, a lorica is “gleichsam wie eine aus magischen Worten geschmiedete Brünne [, die] den Schutz des Beters gegen Angriffe des Bösen gewährleistet”¹¹ (‘like a shirt of mail, forged of magic words, warranting the supplicant’s protection against attacks by the evil’). On that score, the materiality of the words’ denotations cannot be overemphasised. The lexemes of protective armour are supposed to evoke material that saves the physical body in the material world, as well as it protects the soul against invisible attacks. Spoken in terms of metaphorology,¹² the target domain matches with the source domain, in a magic-invisible way. The loricae aim at a materialisation of imagined material, they aim at a magic shield.

Haubrich’s naming of this phenomenon as ‘magic’ seems to be adequate, even more so if Strohschneider’s definition of ‘magic’ is consulted. Strohschneider uses the term referring to a real presence (“Realpräsenz”) of the signified in the signifier, which means a paradoxical identity (“ontische[] Identität und Nicht-Identität”) of sign and thing.¹³ Thus, in the case of hagiographic books which state Mary’s real presence, he defines this dissolution of categorial differences, which cannot be transgressed on the premise of scientific rationality, as ‘magic’.¹⁴ In the case of the loricae, a similar ‘real presence’ can be assumed for the material references of the metaphors – which thereby stop being metaphorical.

The fact that these images were inherited from the Bible might have encouraged this view of words as ‘true things’ – God’s speaking is creating and acting as if word and thing were the same.¹⁵ A German example shows how sacred words were considered material, namely the Münchener Ausfahrtsegen (‘departure blessing of Munich’): unde wil mih hivt gurten / in des heiligen gotts worten […]. Min haupt sei mir hivt stelin / dehainer slaht waffen snide dar in. […] der heilig himeltrvt / der si hivt min halsperch gvt (‘And I will gird myself in God’s holy words today […]. My head be steel today, no kind of weapon may cut into it. […] The holy beloved one of heaven be my good breastplate today’).¹⁶

Against the background of occasional medieval charms that use shield words and sometimes were, written on material, worn as amulets,¹⁷ Reid argues that the loricae were “a conceptual evolution of written words to the status of invisible sentinels”.¹⁸ In contrast to the pagan tradition of “protective amulets and wearable charms”, the Christian “symbolization […] takes the literalness out of the Pagan equation. The iconic physicality of words is no longer required”.¹⁹ Certainly, a lorica does not require being written on material; however, in many cases, the words’ literalness seems to be exactly the point which the conjuring poems aim at; the body parts are commended to this physicality for their protection. Instead of ‘symbolisation’, one might speak with Haubrichs and others of a ‘magic’ material. The materiality of the denotations of the words (instead of their being signs, being symbols) is crucial to this genre, albeit, admittedly, there might be a scale of varying examples and varying sections within the same lorica.

As a complement, loricae also describe the attacks of evil in a material manner (e. g. as ‘darts’, ‘spears’), and here, too, I would question a conventional concept of metaphors. These invisible entities of the ‘spiritual world’ do not seem to be considered very metaphysical in the early Middle Ages, but rather physical as well. At least presentations of the Devil during this time point to this notion. In the Old Saxon Heliand (9^th century), for instance, the Devil must use a heliðhelm[] (helmet which makes invisible) in order not to be seen by Pilatus’ wife,²⁰ apparently due to his very physical nature.

In light of this, a Latin preface of the Lorica of Laidcenn in the manuscript RIA (Dublin) MS 23 P 16 gives another revealing insight into medieval notions of loricae. To expand on the history of that lorica, the account describes how Laidcenn uenit ab eo in insolam Hiberniam; transtulit et portauit super altare sancti Patricii episcopi, saluos nos facere (“came from him to the island of Ireland; he brought it [the lorica] over and placed it upon the altar of Saint Patrick the Bishop, to make us whole”)²¹. Herren reads this as a clear “emphasis […] on the lorica as an object”.²² He assumes:

“[T]he lorica […] was venerated not only as the text of an efficacious prayer, but also as a sacred relic – one worthy, at least according to legend, of the altar of the church at Armagh. Is it possible that loricae were meant not only to be recited, but also to be carried on the person, or even worn (perhaps like a scapular with a long inscription)? Such a hypothesis would place an entirely new construction on St. Paul’s words: induti loricam fidei!”²³

Herren’s hypothesis is also based on his idea that pagan amulets served as generic models for the loricae,²⁴ which, however plausible, cannot be proved. Nevertheless, according to its contemporary preface, this protecting lorica indeed seems to have been conceived an object like a sacred relic given to the altar – similar to weapons and armours of saints.²⁵ At this point, a decision cannot be made whether this ‘thingness’ is tied to the materiality of parchment and letters (see below, chapter 4). It seems clear, however, that this verbal object was deemed special because of its content in the first place. One might conclude that the content made the poem a sacred object; the words that conjure a material-like shield are not words as signs, but object-like as well – signifier and signified coincide. This matches Strohschneider’s definition of ‘magic’.

There is another characteristic of loricae which can be defined as ‘magic’ as well as ‘physical’: their special sound. Many of these prayers – Sancte sator at the extreme end – are structured by devices of repetition (e.g., anaphors, alliteration, rhymes, rhythm),²⁶ which draw the attention to the empirical sound side of language. Some (again, Sancte sator in an impressive fashion) contain obscure words,²⁷ which, lacking a meaning, also foreground their sound side. The following analysis of Sancte sator shows how these sound effects contribute to the physical protective function of the shielding terms, and how one medieval writer thought this ‘sound-object’ required a physical reaction.

The Thingness of Sancte Sator

Sancte sator is a Latin work handed down to us in eight manuscripts from England and the continent, the oldest from the early 9^th century.²⁸ It consists of octosyllables with a pronounced trochaic rhythm, a medial caesura in each line, and bisyllabic rhymes. The cluster of these features is also found in some poems by Theodor of Tarsus, Archbishop of Canterbury († 690), who, therefore, is discussed as the author.²⁹

Rhymes are not to be taken for granted in poems of that time.³⁰ Brinkmann assumes that in the early Middle Ages the realm of rhymes is ‘magic language’, for, in his survey, he consistently finds a co-occurrence of rhymes with magical, adjuring, imploring, religious and cultic contents of verses.³¹  “Reim ist eine Erscheinungsform sprachlicher Analogie, die magischer Wirkung dienstbar gemacht wird; mit seiner Hilfe wird Macht ausgeübt.”³² (‘Rhyme is a form of verbal analogy that serves a magic impact; with the aid of it, power is exercised.’) According to Brinkmann, concerning the early Middle Ages, rhyme could be seen as a ‘weapon’ (“Waffe”) against dark forces.³³ Although this thesis is based on his impressive empirical survey, it lacks further reasoning. Scholars who study premodern magic spells also claim that diverse structures of repetitions, such as alliterations, anaphors, rhythm, and rhymes, are constitutive factors of magic effects.³⁴ In this field, Geier tries to explain (following up Lévi-Strauss): The arranged order (“Überstrukturierung”)³⁵ of spell language is meant as an act of interference in chaotic disorder and is supposed to structure and determine physical reality by using analogy.

“Das magische Denken erarbeitet und bastelt sprachliche Überstrukturen, strukturierte Gesamtheiten, um so den unerkannten Determinismus im Ganzen zu manipulieren […]. Die strukturalen Arrangements der Zaubersprache stehen als evokative Äquivalente für den ontologischen Determinismus. […] [Die] Wirksamkeit verdankt sich der Überzeugung, daß das Arrangement der sprachlichen Ordnung genau jener gegenständlichen Ordnung ähnlich ist, die magisch gewünscht, besprochen und supplementiert wird.”³⁶

(‘Magic thinking creates and builds complex linguistic structures, structural unities, in order to manipulate the determinism which is hidden in everything […]. The structural arrangements of incantations, as evocative equivalents, stand for the ontological determinism. […] The effectiveness is based on the notion that the arrangement of the order of words equals the order of objects which is magically wanted, conjured and supplemented.’)

To Geier, the nature of this ‘magic equivalence’ is absolutely non-sensuous and, therefore, far away from any onomatopoetic physique.³⁷ 

Next to these structures of sound repetition (unison, “Gleichklang”), premodern spells show another essential characteristic: ephesia grammata, uncommon words (“Fremdklang”) that obscure meanings and cause enigmatic mystification, are also regular components of ‘magic language’.³⁸ Both features, unison and opacity, have in common that they draw the attention away from the words’ semiotic side to their physical side of sound. Therefore, it seems to be the physicality of language that causes power, next to or even more than its abstract quality of significance. This may be the ‘missing link’ to Geier’s explanation, because how else could spell language have determining effects on the material world, if this language itself does not become material-like? One might disagree with Geier’s thesis that highly structured language has just a non-sensuous relation to the objects in the world. Even though Geier clearly emphasises a lyrical poem’s ‘materiality’ (he uses quotation marks) which the principle of equivalence draws attention to,³⁹ he fails to make use of this thought in his work on incantations where he finds the same principle operating.

If unison and opacity can be seen as techniques of bringing the physical side of language to the fore, and if the physicality of words fits as the missing link in the reasoning why magic words cause an effect on the material world, that leads to a view of sound as material. This view can be fed back on historical notions. In the Middle Ages, sound was thought as something almost material: it is materialior, to speak with Robert Kilwardby (he uses the comparative to express that, in comparison with numbers, sound is ‘more material/physical’).⁴⁰ For sound is not abstract, but empirical, and – according to Bonifatius and Johannes de Muris – it moves in space like material.⁴¹ That might have facilitated the notion that sound can have power in the material world. This consideration is supported by examples from customs – e. g. screams as non-semiotic sounds are believed to have the power to move the Moon, according to Hrabanus Maurus⁴²  – as well as from fictional texts, which show how causalities are understood beyond the limitations of empiricism: in the Middle High German romance Iwein, a strong knight is thrown to the ground by the mere sound of a thunder; in Lanzelet, the mortal danger in the ‘crying swamp’ is explicitly caused by its sound, independent of the heat.⁴³ If sound is seen as a moving, empirical reality, then it may also have power over things in space – a power which is concentrated in verbal phrases that foreground the physical sound side of words.

As indicated previously, Sancte sator does not only use rhymes; it carries all of the ‘magic’ features discussed here to the extreme. Turning to this extraordinary case, I will now outline how these characteristics effect a ‘thingness’ of the poem and how they strengthen the notion that it is a useful tool for material protection.

The poem is cluttered with rhymes – between half-lines (100%) and sometimes between the ends of lines (30 %)⁴⁴ – as well as with alliterations, and it has a pronounced rhythm:

Sancte sator   suffragator
legum lator    largus dator
iure polens    es qui potens
nunc in ethra    firma petra
a quo creta    cuncta freta
[…]
(vv. 1–5)

This ‘network of sound’ gives the impression of a dense, impenetrable ‘thing’ because every syllable is ‘bound’ to the other syllables in a strict, firm and compelling way by having its obligatory place given by rhythm, rhyme and other sound structures. The half-lines as the smallest rhythmic entities are extremely short so that the individual links are the more fixed: without corroding it, it is not possible to penetrate or change this unity by additions or rearrangements of words. It is not necessary to regard this ‘structure thing’ as a material one, but clearly, it presents itself as a unitary object. According to Vollmann’s system, this is perceived in the oral performance as ‘media thingness’.⁴⁵

Next to this high density of elements of unison, the poem also exhibits plenty of uncommon words, sometimes of Greek origin, which have caused great difficulties to medieval and modern translators.⁴⁶ The semantic meaning, altogether, is rather obscure, some phrases cannot be reasonably translated at all.⁴⁷

Thus, both features of ‘magic language’ abound here,⁴⁸ distracting from the words’ semiotic side (or even impeding it) and foregrounding the physical side of sound. The special sound of this poem has often been admired and described as conjuring or magical, e.g. by Müller: “Der Text stellt eine metrische Reihung ‚magischer‘ Wörter dar, wobei es weniger auf den Wortsinn ankommt, als auf Klang und Rhythmus”⁴⁹ (‘The text is a metrical sequence of ‘magical’ words, stressing sound and rhythm instead of the literal sense’) – Müller, as well as Haubrichs,⁵⁰ also joins ‘magic’ with the sound side of words.

The poem starts with an invocation of God and God’s son, reminding the audience of his power as the creator, and then, from verse 10 on, it invokes his protection. This invocation is done via many notions of armour (e.g., et piacla    dira iacla / Trude tetra    tua cetra, v. 12f. ‘and sins, cruel darts, / thrust the odious aside with your shield’; Christi umbo meo lumbo / sit, ut atro   cedat latro / mox sugmento   fraudulento / pater parma   procul arma / arce hostis, vv. 16–20 ‘May Christ’s shield be on my loins so that the robber with his dark *⁵¹ may straightway yield; father, with the shield ward off the enemy’s weapons far away’). In line with the definition presented above, the poem can be called a lorica beyond doubt.⁵² It conjures protecting armour via the lorica-‘metaphors’ which target their material source domain, as shown in chapter 1. However, the special sound of Sancte sator contributes considerably to the material that is needed for the protection of the body in the material world. Its sound could be seen as (quoting Robert Kilwardby) materialior ‘more material’ compared to the sound of texts which focus on their semantic meaning in the first place. With ‘media thingness’ and magic sound structure which is assumed to be material-like (see above), the prayer offers the use of sound as a quasi-material weapon and protection against evil. Both the denotations of the words and the physical sound are supposed to have the effect of armour. To protect humans in the material world, language becomes close to material in two ways: by metaphors which anomalously target their material source domain and by words that intensely foreground their physical sound side.

Doubling the Weapon: the Poem and its Gloss

In Passau, at “some point after 846” A.D.,⁵³ a German scribe wrote the famous Sancte sator along with Old High German translations into the manuscript clm 19410 (Munich BSB), pp. 39–41. These German translations are reasonably called “gloss” in parts of research literature,⁵⁴ although their special shape, as will be shown below, gave also reason to doubt such a categorisation.⁵⁵ One may accept the term not only for pragmatic but also for two other reasons. The first is their formal appearance in the context of the manuscript, as Voetz stresses: “Die […] althochdeutschen Eintragungen zum ‘Carmen ad deum’ unterscheiden sich äußerlich demnach im Ganzen in keiner Weise von den von demselben Schreiber innerhalb der Handschrift vorgenommenen Glossierungen, wie sie sich unter anderem auch auf der dem ‘Carmen ad deum’ unmittelbar vorausgehenden Seite 38 zu verschiedenen alttestamentlichen Büchern finden lassen”⁵⁶ (‘The Old High German entries with the Carmen ad deum do not differ in their appearance in any way from the glosses written by the same scribe in the very same manuscript, as those on books of the Old Testament on page 38 immediately preceding the Carmen ad deum’) – these glosses have always been categorised as ‘contextual glosses’,⁵⁷ and the scribe did not make a difference. Secondly, scholars agree that both text and translation are a copy of another, now lost manuscript where the translation, at least most likely, had been a regular interlinear gloss (see below). In this sense, being ‘gloss-like’ and probably a former interlinear gloss, the German translation can be called ‘gloss’ as Gretsch, Gneuss⁵⁸ and others do. As it is, clm 19410 contains the only glossed version of Sancte sator that survived.

The Old High German gloss translates every word of the Latin poem. So, at first sight, the gloss seems to strengthen the poem’s semantic side. However, the gloss does not make the obscure phrases clearer (as do, for instance, the Old High German translations by Notker III of St Gallen which explain and interpret obscure images)⁵⁹. It is rather a word-by-word annotation which does not lighten the dark phrases of the poem at all.⁶⁰ So the ‘magic’ mysteriousness of the poem is preserved. The glossator’s errors even contribute to the opacity, e.g. fana demo kamahhot sint alle uuagi […] de fana skeffe forrent plomun (v. 5f.)⁶¹ (‘by Whom are created all waves which carry flowers from the ship’) translate the Latin verses A quo creta cuncta freta / quae aplaustra uerrunt flostra (‘by Whom all seas which ships pass through were created – the calm seas’). That both the original glossator and the scribe of Passau did not hesitate to hand down that translation as an appropriate one is not just a sign of helplessness or carelessness (as some researchers judge)⁶² but demonstrates that they expected and counted on the opacity of the poem rather than on comprehensible semantics. To them, the poem was not supposed to make ordinary sense in detail. This contemporary reception strengthens the notion that opacity may have caused the special impact and popularity of this verbal object.

Fig. 1: Example of an interlinear gloss: Lorica of Laidcenn with Anglo-Saxon (Kentish) gloss of the 10th century, Cambridge, University Library, MS L 1.1.10, fol. 43. Online.
Copyright: University of Cambridge.

The scribe of Passau who copied the poem and the German gloss presumably made a significant change vis-à-vis the master copy. According to the majority of researchers,⁶³ it is most likely that the master copy presented an interlinear gloss, written between the lines of the Latin poem. The scribe of Passau then transformed it into an ‘intra-linear gloss’, alternating with the Latin phrases in the same line (see fig. 2); it attained the shape of a contextual gloss,⁶⁴ which is the form of gloss the scribe uses throughout the manuscript.⁶⁵

Fig. 2: Extract from Sancte Sator with the gloss inserted into the lines (Latin and Old High German alternating within the line):
Photo: BSB Clm 19410, fol. 39, online.

This is a rare phenomenon. Usually, contextual glosses are found in glossaries and commentaries, not in literary works.⁶⁶ The reason is quite obvious: this form of gloss impedes the reading process, it disturbs the unity of the main text which, in case of an interlinear or a marginal gloss, is visually intact (as, e.g., above the glossed Lorica of Laidcenn)⁶⁷. The scribe of Passau disrupts the poem verse by verse with the gloss. With regard to the quasi-magic quality and physicality of this special sound-object, this glossing is a remarkable interference in its integrity and impact. The scribe must have known that to other readers of the manuscript (apparently it was brought to Ilmina, Ilmmünster, in the same century)⁶⁸ the poem could be unknown; his contextual gloss would make it difficult to get access to the poem then. Moreover, the gloss presents itself in the fashion of a word-by-word-annotation, as is typical of interlinear glosses which were read in a vertical direction, each interpretament targeting the lexemes of the primary text.⁶⁹ Therefore, the scholars’ thesis about the master copy of the gloss of Passau seems to be justified. Often lacking a reasonable word order and syntax,⁷⁰ it is unlikely that this gloss was meant to be read as a syntagma on its own. Being integrated into the lines, the interpretaments’ affiliation to a particular Latin lemma is nevertheless loosened; the German words are to be read sequentially as if they, too, were coherent verses, although frequently they have hardly any syntactical coherence.⁷¹ This is the more noticeable in the main section of the poem (from verse 4 on) when the gloss alternates with the primary text every full verse. Being neither a syntagma, nor a single vocabulary, these glosses differ significantly from other contextual glosses, which refer to a single word (lemma)⁷² or, occasionally, offer syntactically coherent translations.⁷³

To the scribe of Passau, perhaps that very quality – the opacity of a sequence of somehow conjuring words – made the gloss so equal to the lorica’s Latin verses that it could seem almost of equal value. Along the same lines, some modern scholars have perhaps overemphasised the rhythm and alliterations of this German ‘text’ in order to treat it as a poetic work as well.⁷⁴ Such views of vernacular glosses have been criticised, for glosses are supposed to require the quality of textuality only in affiliation with the primary text, not per se.⁷⁵ Especially in its shape of a contextual gloss, the work of Passau cannot be continuously read as a work of art because of its regular interruptions by the Latin phrases. In the following, the assumption that the gloss of Sancte sator was not meant to be a self-reliant text will be confirmed; it supplements and interacts with the poem. In this function, however, its own beauty of sound and opacity might have been seen as quite appropriate.

There is one difference between an interlinear and a contextual gloss which has not yet caught much attention in research. It is known that, in the early Middle Ages, reading was not a quiet doing; usually they read aloud or ‘murmured’.⁷⁶ It can be assumed that an interlinear gloss, written in small letters (sometimes only carved without ink) above the text which is to be read, was only meant as a visual help and was not articulated. Entering the same line of the ‘text proper’ and getting equal size, the contextual gloss is included in the linear (sonorous) reading process.

With this in mind, the purpose of the manuscript is also of importance. Wisniewski thinks of a teacher reciting the Latin verses as well as the German words between them, to make his students understand;⁷⁷ this is one possible usage. The manuscript clm 19410 has been defined as an episcopal book for studying;⁷⁸ it contains glossaries, for example, i.e. lists of Latin vocabularies with Old High German translations as contextual glosses (pp. 24, 32f., 36–38, and 58–60);⁷⁹ texts with stress marks for recitation (e.g. pp. 51–53);⁸⁰ and, directly following Sancte sator, basic knowledge about dactyl and trochee as well as Latin set phrases for letters, skills expected from priest examinees (p. 41).⁸¹ The gloss form used for glossaries (contextual gloss) connects the poem with its context in the manuscript. The manuscript’s glossaries, e.g. about “parts of the body, ultimately from Isidore” and “utensils and implements in household and agriculture”⁸² on the same page, are not visually distinguished from each other, they lack any headings and are not collated alphabetically as for example in the Abrogans. Therefore, these lists were rather unsuitable as a dictionary for consulting ad hoc; the monks must have used them for extending their vocabulary,⁸³ for learning by heart (a common way to study for exams and to prepare lessons).⁸⁴ Hellgardt and Rädle suppose that the Latin poem Sancte sator once was generated in school contexts for practising special and poetic vocabularies, e.g. words from the Vulgate and from Vergil’s Aeneis, which are found in the poem.⁸⁵ Indeed, Sancte sator could have been seen as an excellent mnemonic instance. Here one could learn vocabulary in a rhymed and rhythmic order. The function of extending vocabulary is a fortiori the aim of the glossed version. It is likely that the German gloss was not a silent help (in contrast to an interlinear gloss which does not disrupt the reading process); its placing within the lines in the context of vocabulary lists suggest that, at least during the learning phase, the gloss was meant to be spoken or whispered aloud or ‘heard by the inner ear’, but not to be skipped in favour of the Latin verses. The purpose of advanced Latin learning does not mean that the content of the poem was neglected; on the contrary, the belief in its spiritual beneficing as well as its fame (it was part of the devotional book Enchiridion compiled by Alcuin)⁸⁶ made it worth learning. Studying and learning by heart was a crucial part of the spiritual life in monasteries.⁸⁷

In this context, the scribe must have been aware that he determined how the learner and/or teacher preceded and what they, repeatedly, articulated. The following analysis substantiates this and, eventually, shows that the scribe prudently made use of special potentials given by the form of a contextual gloss: he makes it join the poem not only as a commentary but as an additional – and even active – tool supporting the poem’s function in a way which an interlinear gloss would not have facilitated.

It is significant that the intervals of the glosses differ throughout the poem. At the beginning, there are short intervals: every half-line of the Latin poem is followed by the translation, separated by a small dot: Sancte sator ⸳ uuiho fater ⸳ suffragator ⸳ helfari ⸳ / legum lator ⸳ eono sprehho ⸳ largus dator ⸳ milter kepo ⸳ / Iure pollens ⸳ pi rehte uuasanti ⸳ es qui potens ⸳ (vv. 1–3). The German phrases are interlocked with the poem in such a close-meshed way that the Latin rhymes embrace, and thus enclose, the German words as if they were part of the verses. The technique equals a close-meshed chain. However, the interruptions within the Latin verses would endanger the solid ‘magic sound’ of the poem if it were not for the particular four-time-rhyme (-ator) at the poem’s beginning, generating an overall sound unity. Thus, the interpolated German words can become part of the armour poem the learning monk repeatedly articulates, which would not have been possible with an interlinear or a marginal gloss.

After the Latin four-time-rhyme, the gloss disrupts the ‘magic sound’ in a more precarious way (Iure pollens ⸳ pi rehte uuasanti ⸳ es qui potens ⸳ du pist der mahtigo ⸳  / nunc in ethra […], v. 3f.). To avoid (so it seems) these disturbances of the Latin rhymes, the glossing technique then changes. From verse 4 on, the intervals of the gloss are longer – the Latin rhyming couplets are left unseparated: nunc in ethra    firma petra ⸳ nu in himilie fester stein ⸳ / A quo creta    cuncta freta ⸳ fana demo kamahhot sint alle uuagi ⸳ / quae aplaustra    uerrunt flostra ⸳ de fana skeffe forrent plomun (vv. 4–6), and so on.  The ‘magic sound’ of the Latin rhyming couplets is preserved now while the interlinkage of the German phrases with the poem is slightly looser but still evident.

That this structural change takes place in verse 4 (instead of verse 3) does not seem to be accidental; it coheres with the meaning of this verse. Whereas the beginning of the poem calls God a father and helper who is mighty, generous and just, verse 4 contains the first thing metaphor of the poem, and that is a metaphor which describes maximal material solidness and stability (a ‘firm rock’): nunc in ethra   firma petra. Like the firm object of the semantic side, the sound of the rhyming couplet is left intact as a firm integral unit. To be more precise: the verse starts with the spiritual world (ethra ‘heaven’) and connects it with the picture of most solid material. This is exactly the poem’s point of transition between the human-like or abstract representations of the spiritual and the thing metaphors which dominate the text from now on and become essential for this lorica and its effect of protective ability (by ‘magic’ materialisation). This quasi-material ability is apparently linked to the sound effect of the rhyming couplets because these become undisturbed units of sound once the thing metaphors emerge. The material solidness of the content matches the solidness of the sound (which can be considered as material, see chapter 2).

The principle of alternating text and gloss every full line is then consistently carried through in the whole text. It is varied only one time, and that is a meaningful exception. In verse 23, the gloss suddenly interrupts the Latin verse in the cesura:

Tunc deinceps ⸳ denne frammort ⸳ Trux et anceps ⸳ / catapulta    cedat multa ⸳ ungahiuri ⸳ enti zuifoli⁸⁸ ⸳ allaz sper⁸⁹ snidit⁹⁰ managiu ⸳ (v. 23f.).
(Latin text: ‘then thereon may the fierce and double-faced one / yield to many a missile, or: abandon the missile’)⁹¹

Verse 23 describes the enemy of this imagined war, the Devil or evil, for the last time in the poem, at the very (desired) moment of his final defeat.⁹² This very verse (Tunc de-inceps   Trux et anceps) is an extraordinary piece of art with its interlocking rhymes, alliterations and sound correspondences (Tunc–trux; mirroring de–et; inceps–anceps). Just here, an interruption disturbs the sound-unity, which therefore does not have a chance to take effect. The gloss interferes and breaks the ‘magic’ powerful sound impact of this verse describing the evil. It seems as if the scribe did not want to grant the Devil such a powerful sound (and he makes sure that a teacher or ‘murmuring’ learner breaks the verse up halfway, just in case the sound had power). Thus, the gloss damages the powerful sound of the Devil’s emergence just as the content of the next verse of the poem describes his defeat. The technique of glossing imitates exactly the semantic meaning of the gloss: cedat is translated with snidit (‘cuts’), and this is what the gloss itself physically performs. On the level of the word-by-word-gloss there is hardly any semantic coherence (ungahiuri enti zuifoli are uninflected annotations, managiu just literally corresponds to multa), but on the level of the poem’s sound, verse and gloss structure, it is clear: the enemy’s power should be ‘cut’ by the ‘spear’ – (gloss:) sper snidit – just as his powerful sound, the physically sonorous verse, is ‘cut’ into two pieces (by the gloss) und thus ‘disarmed’. In this way, the gloss itself becomes the ‘cutting spear’ on the level of the materiality of the text, a weapon just as the lorica itself was meant to be. Through the way in which gloss and poem interact, the armour built against evil becomes even more powerful.

Conclusion

Sancte sator achieves its apotropaic power through invoking the material side of its armour metaphors (neglecting their semiotic structure and figurative nature). It is meant as a lorica, a shield against evil, and intensely uses features (rhymes, alliterations, strict rhythm, uncommon words, obscure phrases) which stress the (in medieval view material-like) physical sound of language instead of the semiotics. In this way, on several levels, Sancte sator gains ‘thingness’ as a physical shield and weapon. The presumed lost Old High German interlinear gloss annotated the poem word by word, without enlightening the enigmatic opacity, and left it unchanged for the (sonorous) reading process as an intact object of sound. The scribe of clm 19410 (Munich BSB) found this gloss, lacking transparent coherence and meaning just as the Latin poem. He joined the two of them into the same lines, creating one verbal object for meditating and learning purposes. The close-meshed interlacing slightly disturbs the powerful sound. However, the intact sound-object (untouched rhyming couplets) unfolds just at the moment when the invisible (ethra) is connected with the first thing metaphor of the poem, whose source domain is not by chance firm and unbreakable material (firma petra). Thus, the material meaning of the words is linked to the special sound effect of the poem and its ‘magic’ ability to protect. When it comes to the final defeat of the enemy, ‘his’ rhyming couplet (his material sound-power) is ‘cut up’ by the prematurely intervening gloss which imitates its own content, a cutting weapon against evil (sper snidit). In its interaction with the lorica, the gloss becomes a sper at the critical moment of the imagined fight.

In conclusion, diverse dimensions of materiality are essential for the Sancte sator in the manuscript clm 19410: the basic physical side of language (sound), the physical meaning of the words (source domain of thing metaphors; the real presence of the signified in the signifier), and, on both levels, the physical interaction of gloss and poem (the sound-object is not left untouched, but carefully joined and even ‘cut’ by the spear-imitating gloss). This finally leads to the question as to what extent the visual materiality of the manuscript is determinant as well.

We might, therefore, take a broader perspective and ask whether and how medieval glossators treated the texts they commented on, and even the glosses, as material objects. Famous examples of Glossae Ordinariae (extensive marginal comments on and interlinear glosses between the lines of the main text) present aesthetic visual constructions; different colours and sizes of characters as well as elaborate decorations and frame features of the written emphasise the materiality and its value (a value which is also attributed to the glosses)⁹³. Significantly, these examples are to be dated in later centuries when, according to Ivan Illich, a book page transformed from a ‘score for murmurers’ to a ‘visual construct for thinkers’.⁹⁴ However, for the early Middle Ages the materiality of scriptural artefacts has often been stressed in other contexts as well, especially in the case of spells which were effective as characters,⁹⁵ e.g. as amulets, and also in the case of the Lorica of Laidcenn (see above, chapter 1). Hartung and others even guess that in the early Middle Ages, when only a few people could write and the alphabet was used especially for the tradition of sacred texts, the mere characters in their material presence were considered magic entities.⁹⁶ The gloss of Passau, visually unified with the lorica poem, might be seen as completing the ‘magically’ effective artefact of letters. However, the owners and users of this manuscript – advanced scholars and learners – seem to have had a different view of this codex. Features which bring the materiality of the written to the fore are rare in the whole part of the manuscript; it virtually lacks any differentiating marks, headings, character size distinctions, non-verbal signs or decoration. The gloss of Sancte sator contrasts the presumed former interlinear gloss of the master copy which left the ‘text proper’ visually intact and treated it as a two-dimensional ‘artefact proper’ or textus (‘fabric, text’) into which the gloss was woven as a visibly distinct, thinner ‘thread’. In the case of the Passau version, given its purpose in the monastic school context (learning, murmuring, reciting, praying), one might ask if the construction of the text suggests neglect of the written in favour of other dimensions of materiality: the palpable sound and the shield meaning. So, in order to approach the relation of glosses and primary texts, I would hesitate to generalise a visual materiality focus of a ‘manuscript culture’ because that focus too easily distracts from other forms of the materiality of words in a culture which is, to a great extent also in monasteries, primarily ‘vocal’.

Abbreviations

AhdGr: see ‘Reference Works’.

DVjs: Deutsche Vierteljahrsschrift für Literaturwissenschaft und Geistesgeschichte. 1923 ff.

MGH: Monumenta Germaniae Historica. https://www.dmgh.de/

NG: see ‘Reference Works’.

PBB: Beiträge zur Geschichte der deutschen Sprache und Literatur. 1874 ff.

PL: Patrologia Latina. Ed. by Jacques-Paul Migne. 161 vols. Paris 1857–1866.

VL: Verfasserlexikon. Die deutsche Literatur des Mittelalters. Ed. by Kurt Ruh, 14 vols. Berlin, 1977–2008.

ZfdA: Zeitschrift für deutsches Altertum und deutsche Literatur. 1841 ff.

Sources / Text Editions

Robert Kilwardby: De ortu scientiarum. Ed. by Albert G. Judy. Oxford 1976.

Bonifatius: Letter 115. MGH Epist. sel 1.

Johannes de Muris: Notitia artis musicae. Ed. by Ulrich Michels, o. O. 1972.

Behagel, Otto (ed.): Heliand und Genesis. 10^th ed. revised by Burkhard Taeger. Tübingen 1996.

Hartmann von Aue: Gregorius. Der Arme Heinrich. Iwein. Ed. by Volker Mertens and Walter Haug. Frankfurt/M. 2004.

Hrabanus Maurus: Homiliae de festis praecipuis. In: PL 110, p. 78f.

Lorica of Laidcenn. Ed. in: Singer, Charles: The Lorica of Gildas the Briton (? 547). A Magico-medical Text containing an Anatomical Vocabulary. In: Proceedings of the Royal Society of Medicine 1919 (12), pp. 123–144, here pp. 136–140.

Lorica of Laidcenn. Ed. in: Grattan, John Henry Grafton / Singer, Charles (ed.): Anglo-Saxon Magic and Medicine. Illustrated specially from the Semi-Pagan Text “Lacnunga”. London 1952, pp. 130–147.

Sancte Sator. Ed. in: Hellgardt, Ernst: Das lateinisch-althochdeutsche Reimgebet ‚Sancte Sator‘ (sog. ‚Carmen ad Deum‘) Theodor von Tarsus / Canterbury zugeschrieben. In: ZfdA 2008 (137), pp. 1–27, here pp. 9–14.

Ulrich von Zatzikhoven: Lanzelet. Text – Übersetzung – Kommentar. Ed. by Florian Kragl. Berlin 2009.

Reference Works

AhdGr: Braune, Wilhelm: Althochdeutsche Grammatik I. Laut- und Formenlehre. Ed. by Ingo Reiffenstein. Berlin 2004.

NG: Der neue Georges. Ausführliches lateinisch-deutsches Handwörterbuch. Aus den Quellen zusammengetragen und mit besonderer Bezugnahme auf Synonymik und Antiquitäten unter Berücksichtigung der besten Hilfsmittel ausgearbeitet von Karl-Ernst Georges. Ed. by Thomas Baier. 2 vols. Darmstadt 2013.

Prinz, Otto/Stroux, Johannes/Schneider, Johannes/Lehmann, Paul (ed.): Mittellateinisches Wörterbuch bis zum ausgehenden 13. Jahrhundert. Bayerische Akademie der Wissenschaften; Deutsche Akademie der Wissenschaften; Berlin-Brandenburgische Akademie der Wissenschaften. München 1999.

1. This title was invented by John A. Giles, who edited the Text in 1850, see Gretsch/Gneuss 2005, p. 13.
2. The term is coined by Wieland 1983, p. 7, who distinguishes between ‘text proper’ and ‘gloss’. His rather loose definition of a gloss as “anything on a page which is not text proper, but which is intended to comment on the text” includes contextual glosses as well. In this regard it conforms to the definition by Henkel 2007, p. 727: “Glosse[:] Instrument der Texterschließung, bezogen auf das Einzelwort oder einen umfänglicheren Textzusammenhang. […] Die Glosse kann zwischen den Zeilen stehen (Interlinearglosse; Interlinearversion), innerhalb der Zeile (Kontextglosse) oder am Rand (Marginalglosse)” (‘gloss: instrument for exploring texts, referring to a single word or a more extensive section. The gloss can be placed between the lines [interlinear gloss; interlinear version], within the line [contextual gloss] or at the margin [marginal gloss]’). Restricting the term ‘gloss’ to those metatexts having visually a paratextual position on the page makes contextual glosses an exception (cf. Bergmann 2020, p. 125; Stricker 2009, pp. 20–32; see below, note 64).
3. Cf. Hill 1981, esp. pp. 264-266; Reid 2002, pp. 142f., 151f.
4. Cf. Hill 1981, pp. 259–266; Reid 2002, p. 141.
5. Reid 2002, p. 141; cf. Stifter 2007, p. 507f.
6. Haubrichs 1995, p. 220; Brinkmann 1960, p. 68.
7. A comprehensive overview of these biblical passages is given by Stifter 2007, p. 506f.
8. Quotations (with translation from paragraph 16 on) from: Singer 1919, p. 138.
9. Quotations of Sancte sator from: Hellgardt 2008, p. 10.
10. Hill 1981, p. 266; see also p. 267, and Reid 2002, pp. 143, 145, 152.
11. Haubrichs 1995, p. 220.
12. For a brief introduction into the technical terms in use see Kohl 2007, pp. 30–37.
13. Strohschneider 2004, p. 263, highlighting ibid.
14. Strohschneider 2004, p. 263; see also Haeseli 2011, pp. 189–194, 201.
15. Cf. Schulz 2000, p. 186f.
16. BSB, Munich, Clm 23374, fol. 15v f. (quotations 16r), with conjecture (himeltrvt instead of the manuscript’s himelbrvt) according to Holzmann 2001, p. 270f. Cf. also Schulz 2000, p. 187.
17. Cf. Reid 2002, p. 152.
18. Reid 2002, p. 153.
19. Reid 2002, p. 153.
20. Heliand, v. 5452 (Behagel [ed.] 1996, p. 192).
21. Latin text and translation quoted from Singer 1919, p. 137.
22. Herren 1987, p. 30.
23. Herren 1987, p. 31.
24. Herren 1987, p. 31.
25. E.g. the lance said to be either St Mauritius’ (patron saint of the armourers and soldiers) or St Longinus’ (the Roman captain who wounded Christ’s body) and then emperor Constantine’s; this lance was believed to have saved the city of Modena from the Hungarians in 899 and to have helped Heinrich I and Otto I to win battles in the 10th century (cf. Wolf 2005, pp. 25–44; Worm 2000, pp. 183–197; Suckale-Redlefsen 1987, pp. 34–36). Other examples are St Wenzel’s helmet, mail shirt, sword and lance (cf. Podlaha/Šittler 1903, pp. 9–16; Naegle 1928, pp. 13, 135–137; Machilek 2018, p. 65) and St George’s lance tip (Braunfels-Esche 1976, p. 78).
26. Cf. Brinkmann 1960, pp. 70, 72.
27. “Some loricae tend to use unusual terminology culled from the ‘holy’ learned languages (Hebrew and Greek); this is the case of the Lorica of Laidcenn” (Reid 2002, p. 141).
28. Cf. Hellgardt 2008, p. 5f.; Gretsch/Gneuss 2005, pp. 9–11.
29. Ascription by Lapidge 1996, pp. 225–239; to other scholars, Lapidge’s thesis is very likely, see e.g. Hellgardt 2013, p. 67; Hellgardt 2008, p. 3; Gretsch/Gneuss 2005, pp. 9, 25, 33–37.
30. According to Brinkmann 1960, pp. 67–76, it was the Irish tradition that gave the critical impulse for rhyming Latin poetry to the rest of Europe in the early Middle Ages.
31. Brinkmann 1960, pp. 69, 64.
32. Brinkmann 1960, p. 63.
33. Brinkmann 1960, p. 71.
34. E.g. Feulner 1999, p. 131; Geier 1982;  Beck 2003, p. 283f., 346f.; Steinhoff, col. 417.
35. Geier 1982 identifies the principle of spells with Roman Jakobson’s principle of equivalence, e.g. p. 377.
36. Geier 1982, pp. 381, 383 (highlighting ibid.); see also pp. 380–382.
37. Cf. Geier 1982, p. 384, following up Walter Benjamin.
38. Cf. Brinkmann 1960, pp. 72, 75 (quotations); Schwab 1995, p. 264; Beck 2003, pp. 313, 347; Geier 1982, p. 376.
39. Geier 1979, e.g. p. 57.
40. Robert Kilwardby: De ortu scientiarum (Judy [ed.] 1976, p. 57, Nr. 143). – Just as impressions of visible material objects seen by the eye, sound, according to medieval theories, is finally perceived by the sensus interior, which produces a synesthetic image in the intellect’s chambers of imaginatio, phantasia and memoria. This is why medieval texts can present perception of women’s beauty through perceiving sound, as e.g. Schneider 2020, p. 165f., shows (with further research literature).
41. Cf. Bonifatius’ Letter 115, MGH Epist.sel. 1; Johannes de Muris: Notitia artis musicae II, III, 2 ( Michels [ed.] 1972, p. 71).
42. Hrabanus Maurus, Homiliae de festis praecipuis, PL 110 78f.
43. Hartmann von Aue: Iwein (ed. Mertens/Haug [eds.] 2004), vv. 650–652; Ulrich von Zatzikhoven: Lanzelet (Kragl [ed.] 2009), vv. 7062–7065. – Cf. Marshall/Stellmann 2019.
44. Cf. Schröbler 1957, p. 31.
45. See Vollmann 2020 (in this issue of MEMO).
46. See e.g. Schönbach 1898, pp. 113–119; Rädle 1978, col. 1176; Gretsch/Gneuss 2005, p. 21; Müller 2007, p. 367; Hellgardt 2013, p. 69f.
47. E.g. v. 20f.: uti costis / immo corde; v. 29: Sicque ab eo   me ab eo. Cf. Gretsch/Gneuss 2005, p. 14: “All in all, considering the extremely difficult language of Sancte sator, it seems remarkable that it has not suffered more in the course of its transmission”.
48. Cf. Brinkmann 1960, p. 72.
49. Müller 2007, p. 367.
50. Cf. Haubrichs 1995, p. 220: “[Das] ‘Panzerlied’ sollte – gleichsam wie eine aus magischen Worten geschmiedete Brünne – den Schutz des Beters gegen Angriffe des Bösen gewährleisten. Daher erklärt sich die eindringliche, in ihrem steten, fast monotonen Gleichlauf von Alliteration und Binnenreim gestützte litaneiartige Form” (‘The ‘armour poem’ – like a shirt of mail, forged of magic words – should warrant the supplicant’s protection against attacks by the evil. This explains its forceful form, which due to its monotony of alliterations and rhymes is litany-like’). Haubrichs and Müller do not clarify, however, why the protective function actually should explain this form.
51. I skip sugmento which is highly disputed; Lapidge 1996, p. 243, thinks of a misspelling of augmento and translates “with his black, deceptive growth” (p. 241); Schönbach 1898, p. 114f., takes it for sagmentum (the form sugmentum is documented by another gloss), stemming from the Greek word σάγμα, and translates “in seinem dunklen trügerischen kleide” (‘in his dark deceitful vesture’) (p. 114).
52. Cf. also Rädle 1978, col. 1176; Haubrichs 1995, p. 220.
53. Gretsch/Gneuss 2005, p. 22. Cf. Hellgardt 2013 pp. 1249–1251 (Nr. 660).
54. E.g. Gretsch/Gneuss 2005, p. 16, 18, 21 passim; Hellgardt 2008, p. 2 (“Glossierung”); see also below, note 64.
55. In the BStK Datenbank der althochdeutschen und altsächsischen Glossenhandschriften, the text is not listed: https://glossen.germ-ling.uni-bamberg.de/texts and https://glossen.germ-ling.uni-bamberg.de/glossed_contents/12812.
56. Voetz 2009, p. 893f.
57. Cf. BStK Online: https://glossen.germ-ling.uni-bamberg.de/glossed_contents/12812 with further research literature.
58. Gretsch/Gneuß 2005
59. See e.g. Hehle 2003, p. 187–190, 196–198, 202; Hehle 2002, pp. 116, 120–123, 125–127 and passim.
60. Cf. Rädle 1978, col. 1176; Schönbach 1898, p. 117.
61. Quotations (Latin text and OHG gloss) again from Hellgardt 2008, p. 10. 
62. Even if scholars do not think as little of the glossator’s skills as Baesecke, Koegel and others (for an overview see Gretsch/Gneuss 2005, p. 21, Schönbach 1898, p. 113), they do not consider that the glossator understood exactly and in a deeper sense how the poem worked and very well perceived its opacity, which can be adapted in Old High German.
63. Only Voetz 2009, p. 894, is sceptical about any presumptions. Favourers of an original interlinear gloss are e.g. Hellgardt 2013, p. 67; Gretsch/Gneuss 2005, p. 21 passim; Wisniewski 2003, p. 110; Henkel 1996, p. 70f.; Rädle 1978, col. 1176; Schönbach 1898, p. 117. To Gretsch and Gneuss, this first (lost) Old High German interlinear version is very likely grounded on Latin glosses because there are indications that sometimes “the glossator translated a Latin explanatory gloss to the lemma and not the lemma itself” (Gretsch/Gneuss 2005, p. 22). Some of these errors could even be “possible evidence for lost Anglo-Saxon glosses and annotations” the German glossator had access to (Gretsch/Gneuss 2005, p. 21).
64. Cf. Müller 2007, p. 368: “Beim Abschreiben wurde das Deutsche dann in Form einer Kontextglosse in den lat. Text integriert.” Hellgardt 2008, p. 2 with note 6: “[E]s wechseln innerhalb der Zeile, also intralinear, lat. Text und ahd. Glossierung ab[.] [ibid., note 6:] Man spricht in solchen Fällen, die allerdings selten sind, auch mit einem wenig passenden Terminus von ‘Kontextglossen’; in den Glossenteilen von E [manuscript of Sancte sator] begegnet diese Aufzeichnungsform noch mehrfach”; Henkel 1994, p. 70: “die deutsche Version ist in Art der Kontextglosse in die Zeilen eingefügt”; Haubrichs 1995, p. 220: “‘Carmen ad Deum’, das Zeile um Zeile (in der Manier der Kontextglossen) das aus insularer Tradition sich herleitende […] Schutzgebet […] in die Volkssprache übertrug”; Müller 2003, p. 317: “wie eine Kontextglossierung”. Some of the quotations show that the term ‘contextual gloss’ is met with some reserve in this case; for a general discussion of the term see Bergmann 2020, showing interesting cases, for instance the German vocabularies found within the lines of the Latin Herzog Ernst C: “Den Schreibern […] ist der Status der deutschen Wörter als Erklärung der speziellen lateinischen Vokabeln bewusst gewesen, wie die durchgehende explizite Einbettung mit .i. ‘das heißt’ zeigt. Sie stehen zwar linear im Kontext, wurden aber offensichtlich nicht einfach als Textbestandteile angesehen” (p. 122). The term ‘contextual gloss’ is disputed, though (ibid.). Some call the Old High German Sancte sator material simply “gloss”, e. g. Gretsch/Gneuss 2005, p. 16, 18, 21 passim; because of its originating from an interlinear gloss, this is an adequate term. With regard to its special position within the lines I use the term (in line with the quotations above) ‘contextual gloss’ although this is not a standard instance because contextual glosses are mostly found in glossaries and commentaries (see below, note 66).
65. Cf. Voetz 2009, p. 893f. and BStK Online: https://glossen.germ-ling.uni-bamberg.de/glossed_contents/12812 with further research literature.
66. Cf. Bergmann 2020,  p. 125f.; Stricker 2009, p. 24f.; Hellgardt 2008, p. 2 note 6; https://glossen.germ-ling.uni-bamberg.de/abbreviations/310. The contextual glosses in Herzog Ernst C show that there are exceptions to this rule (cf. Bergmann 2020, p. 122).
67. As it is the case in the manuscript of Cambridge (see Fig. 1), other vernacular glosses also translate each word of the Lorica of Laidcenn above the lines in the manuscripts Dublin, Royal Irish Academy, MS 23 P 16, pp. 241b–242b, and London, British Library, Harley MS 585, f. 152r–152v (Grattan/Singer [eds.] 1952, pp. 130–147).
68. Cf. Brunhölzl 2000, p. 32f.; Hellgardt 2008, p. 5; Bergmann/Stricker 2005, p. 1249; Bischoff 1971, p. 125f.
69. See the fundamental work by Henkel 1996 (about the Carmen ad deum pp. 65f., 70f.); Henkel 2009, p. 495 and passim.
70. E.g. v. 15: in hoc sexu   carnis nexu (‘in this sex in the flesh’s bondage’) . in desemo heite fleisc kapunta (‘in this shape flesh [nominative!] bound’); v. 28: dicam deo   gratis geo (‘I may say I give/pour out thanks to God’) . ih quidu . cote dancha . toon (‘I say to God thanks to do’ [infinitive! The first person is tom, see v. 14 and AhdGr, p. 310]). So just annotating lemma after lemma, the glossator sometimes chose the uninflected forms (nominative, infinitive) and strictly adhered to the poem’s word order (see also v. 17!).
71. See above, note 60 and note 70.
72. Examples in the very same manuscript, pp. 58–60; see also the Glosses of Kassel, e.g. Schmid 2009, p. 1080, and the glosses in Herzog Ernst C, cf. Bergmann 2020, p. 122.
73. E.g. the Glosses of Reichenau Rb. For coherent syntagms in glosses see Schmid 2009, pp. 1079–1088; Bergmann 2009, p. 1090.
74. To Sonderegger 2003, p. 101, 108, 124–126, 154, the gloss is poetically designed (p. 108) and impressive by means of its rhythmic impetus (p. 124f.). Cf. also Haubrichs 1995, p. 220 (‘contrafactum’). This view has been criticised by Henkel 1996, p. 70f., and Murdoch 2008, p. 10.
75. Cf. Henkel 1996, esp. p. 66–68, 70f.; Schiegg 2013, pp. 48, 65; Bergmann 2009, pp. 1090–1093; Murdoch 2008, p. 10.
76. Cf. Schiegg 2013, pp. 26, 143, 170; Green 1996, pp. 15–17, 31–35, 148 f.; Illich 1991, pp. 57–60; Schreiner 1997, pp. 2–6. This also applies to the writing process, see Green 1996, pp. 16, 148.
77. Cf. Wisniewski 2003, p. 110f.
78. Cf. Hellgardt 2008, p. 7f. For the problematic terms ‘classbook’ and ‘library book’ see Schiegg 2013, pp. 125–178; Wich-Reif 2009, p. 661f. To Haubrichs 1995, p. 219f., the poem served metrical lessons and the collected texts of the manuscript were typical of school books; similarly Brunhölzl 2000, pp. 36, 47; Braun 2017, pp. 160–162, 262, 266; Gretsch/Gneuss 2005, p. 34f., also guess that the former copies of Sancte sator, being glossed, were available in Theodore’s and his students’ (e.g. Aldhelm’s) “classroom”.
79. These glosses are edited by Frank 1984, pp. 129–136. For further studies see Bergmann/Stricker 2005, p. 1250f.
80. Cf. Hellgardt 2008, p. 7; Brunhölzl 2000, p. 36.
81. Cf. Schiegg 2013, p. 175.
82. Gretsch/Gneuss 2005, p. 17; see the edition of these glosses by Frank 1984, p. 133f.
83. For the use of biblical and antique glossaries for advanced studying see Wich-Reif 2009, p. 662; also in cases of everyday vocabulary, this is a common purpose of such glossaries, cf. Stricker 2009b, p. 598.
84. Cf. Schiegg 2013, pp. 134, 138, 175f.; Wich-Reif 2009, p. 662; Glauch 2003, p. 209.
85. Cf. Hellgardt 2013, p. 69; Rädle 1978, pp. 117–119; Schönbach 1898, p. 118f.
86. Cf. Hellgardt 2013, p. 67f.; Hellgardt 2008, p. 6f.; Gretsch/Gneuss 2005, pp. 14–16, 36.
87. Cf. Schiegg 2013, pp. 145, 169–178; Illich 1991, pp. 22, 55–65.
88. This translation is likely to have derived from a Latin scholion or gloss to anceps and not from the lemma anceps itself, see Gretsch/Gneuss 2005, p. 23.
89. sper is the translation of catapulta, apparently misled by a Latin scholion on the Greek loanword cata-pulta that the glossator had access to, see Gretsch/Gneuss 2005, p. 25.
90. The glossator apparently understood cedat as a form of caedere, see below note 91; an object is missing because all nouns in the gloss are nominatives.
91. Scholars’ emendation of cedat (all manuscripts, except F: cedit) to cadat (‘may collapse’) (e. g. Gretsch/Gneuss 2005, p. 12f.) is not necessary; Müller’s translation (Müller 2007, pp. 215, 369) – deriving cedat from caedere – with ‘Möge der Grimmige und Trügerische durch viele Wurfgeschosse stürzen’ (‘May the fierce and deceitful one fall through many missiles’) is not possible either, because caedere ‘stürzen’ is always a transitive verb (‘jemanden stürzen’ = ‘to make somebody fall’). So I agree with Wipf 1992, p. 113, that cedat is a form of cedere ‘yield’; however, as Müller, I do not take catapulta as nominative, but as ablative (either ablativus instrumentalis or dependent from cedere ‘yield to’, ‘relinquish / abandon / give up [possession]’, ‘cease from’, cf. NG, vol. 1, col. 817; Prinz et al. 1999, vol. 2, col. 424). multa can also be used adverbially (Lapidge 1996, p. 241, 243: ‘straightway’). Or is it (cf. Wipf 1992, p. 113) multa f. ‘penalty, penitence’ (‘may he relinquish the missile as a penalty / due to my penitence’)?
92. Following, there are just prayers to (probably) Mary for blessings and a prospect of thanks to Christ.
93. Cf. Schiegg 2013, pp. 162–164, 266–271. For examples of Glossae Ordinariae see e.g. Cologny, Fondation Martin Bodmer, Cod. Bodmer 31, f. 2ff.; Munich, BSB Clm 28163 (see e.g. fol. 124r).
94. Cf. Illich 1991, p. 8.
95. Cf. e.g. Haeseli 2011, pp. 49–59, 105f., 165–167, 187, 200–202; Hartung 1993, pp. 109, 113–120, 124f.
96. Cf. Hartung 1993, pp. 113–120, 124f.; Schulz 2000, p. 186f.