• 1 See for instance Sandler 1999, 19; Solopova 2013, xvi; Van der Horst 1996, 55. NB: This study draws (...)
• 2 This type of manuscript has often been referred to as a biblical Psalter by scholars. See for insta (...)
• 3 On additional materials see for instance Harper 1991, 312; and Sandler 1999, 18–19.
• 4 Solopova lists different “features which suggest that a psalter was made for a lay rather than reli (...)

1Psalter manuscripts were among the most frequently illuminated books of the Middle Ages.1 This is especially true for manuscripts produced in the twelfth to fourteenth centuries (Sandler 1999, 19). It is therefore hardly surprising that they have been and continue to be at the centre of art historical interest. It is not only the illuminations themselves, but also the significance of the medieval Psalter manuscripts in various areas of daily life that has recurrently been the focus of scholarly interest. Psalter manuscripts were made for different purposes and owned by both clergy and laity. One type of Psalter manuscript, thought to have been made mainly for study, contained the 150 psalms in one or multiple Latin versions or in different languages. Glosses could also be added.2 Most of the surviving Psalter manuscripts belong to a different type, namely so-called liturgical Psalter manuscripts. The core content of these manuscripts is the biblical Book of Psalms, but they include additional devotional material as well. Examples are an ecclesiastical calendar that precedes the psalms and biblical canticles from the Old and New Testaments, collects or a litany of the saints that follow the last psalm.3 “Liturgical” implies a use within mass or devotional practices, but the term has also been used for manuscripts owned by and made for laypersons.4 The term does not so much depend on how the manuscripts were actually used, but rather on what they contain and how this content is structured.

• 5 For other elements of the Divine Office, see Harper 1991, 75–86. See also Dyer 2012, 664–70 and Hug (...)
• 6 The distribution of the psalms is relatively complex and therefore we will only refer here to the t (...)

2Liturgical Psalter manuscripts are closely linked to the Divine Office (or Liturgy of the Hours). The Divine Office consists primarily of the daily recitation of the Psalms at eight fixed times (the so-called hours) throughout the day and night. All 150 psalms were recited over the course of one week, and in addition to the Psalms, canticles were sung, there were scriptural readings, the Lord’s prayer was said, etc.5 How exactly the psalms were distributed over the week depended on whether one followed the monastic or the secular (or non-monastic) cursus. In the monastic cursus, based on the Rule of Saint Benedict from the sixth century, monks were to recite Psalms 20–31 during Matins on Sunday, while the secular cursus roughly covered Psalms 1–20.6 Although liturgical Psalter manuscripts contain more than just the Psalms, they do not include all the texts that were needed for the Divine Office. Thus, in addition to the Psalter manuscript, other books were needed for a complete performance, such as an Antiphonary in which all the antiphons that were sung before and after most psalms were listed. The question of whether and how this liturgical practice affected the visual organisation of the manuscripts is at the centre of the following discussion. It will be argued that the Divine Office can be understood as a cultural frame that is reflected in the visual organisation of the manuscripts.

• 7 The Book of Psalms is a collection of individual shorter texts, i.e. the 150 psalms, and this struc (...)
• 8 Frames are, for example, used to further highlight the three psalms of the threefold division in th (...)
• 9 The term “manuscript architecture” is used in the following when referring to the arrangement of th (...)

3Modern readers are used to structuring paracontent such as headings, page numbers and a table of content. All these things make it possible to access specific sections and to find one’s way around the book. Although page numbers, headings or the like are usually not found in medieval Latin Psalter manuscripts, these books often display an elaborate system of decoration that makes them easier to navigate and which may be related to certain practices of use. Initials of different sizes are a regular feature of Psalter manuscripts and are used to structure the manuscripts.7 In addition, painted frames in the margins of individual pages can allow quick access to the pages highlighted in this way, because they are particularly eye-catching when leafing through the manuscript.8 Some of these varied and manifold possibilities of manuscript architecture are the focus of the following reflections.9 Inspired by Erving Goffman’s Frame Analysis, this essay aims at examining different peculiarities of the manuscripts against the background of the Divine Office. Or, to take up the title of this paper, what follows is an attempt to trace the Divine Office on the pages of various Psalter manuscripts.

• 10 Goffman’s detailed distinction between different types of frames is only of marginal interest for t (...)
• 11 Goffman also introduces the term “keying” that he uses for the “process of transcription” (Goffman (...)

4Frames have long been understood not only in the sense of physical objects that frame, for example, a work of art, but they have also been discussed as cognitive structures of perception or interpretation. Frames in the latter sense were discussed in detail by sociologist Erving Goffman in his Frame Analysis, first published in 1974. They can be considered as schemata of interpretation that enable a person to understand and respond to different social situations. Such a primary framework, in Goffman’s terminology, “allows its user to locate, perceive, identify, and label a seemingly infinite number of concrete occurrences defined in its terms” (Goffman 1986, 21).10 A central concept in Goffman’s frame analysis is the so-called key. He uses the term to describe “the set of conventions by which a given activity […] is transformed into something patterned on this activity but seen by the participants to be something quite else” (43–44).11 The setting of a theatre stage, for instance, provides the main clue to the audience that a fight taking place is just play and no serious real-life situation. In the following, Goffman’s basic idea of frames and keys will be adapted to manuscript studies.

• 12 See, for example, the contributions in Wolf and Bernhart (2006) or Wagner and Conrad (2018). In the (...)

5While the concept outlined above is sociological in origin, it has also been used in psychology and linguistics, in literary studies and also in art history.12 In his highly informative introduction to the volume Framing Borders in Literature and Other Media (2006), Werner Wolf himself clarifies the theoretical basis for the application of frame analysis to literature and other media. Following Wolf, frames can be understood as “basic orientational aids” (Wolf 2006, 5) that help an individual make sense of their experiences and correctly interpret them. Wolf’s comparison with a toolbox clearly illustrates the abstract concepts of frames or frameworks as well as (in Goffman’s terminology) keys and keyings: “If the abstract frames can be described as tools of interpretation, their codings in framings are the (visible or imagined) labels on the tool-box that induce the recipient to choose the correct tools” (26).

• 13 I am indebted to Prof. Dr. Bruno Reudenbach, who pointed this problem out to me. Wolf addressed the (...)

6With regard to the study of manuscripts, Goffman’s frame analysis can contribute to a better understanding of the visual organisation of the pages of Psalter manuscripts as well as of their architecture. In keeping with frame analysis, some structural patterns and conventions can be identified “that organize the manuscript both in terms of its physical makeup and its content, making the object manageable and the contents both accessible and comprehensible” (Wimmer et al. 2015, 7). These structures and conventions correspond to what Goffman called keys (or keying, if he refers to the process of transcription). Wimmer et al. introduced the term “patterns” rather than “keys” (or “keyings”) for these structures and conventions that manifest visually on the pages of manuscripts. To facilitate or enable a specific use of the manuscript, patterns must be recognisable without the reader having to carry out a thorough analysis first. Specific kinds of use that are enabled, facilitated or encouraged by the visual characteristics of the respective manuscripts can be seen as possible cognitive frames. As a simple example, initials always mark the beginning of a text, section or sentence. Strictly speaking, it is a simplification to speak of only one frame: knowing that an object is a book and knowing how to handle a book, for example, are also conditioned by at least one cognitive frame. In this respect, frames cannot be conceived of as isolated structures.13 But, for the sake of clarity, I will speak of single frames in the following.

• 14 This differentiation was also made by Wimmer et al. On the two types of patterns, see in the follow (...)
• 15 This has been shown, for example, for Aristotle manuscripts, which could be used to look up specifi (...)

7There are two categories of patterns that are common in Psalter manuscripts, namely those that concern the content and those that regard the visual organisation of the respective manuscripts.14 Content-related patterns are, for instance, genre-specific (para-)texts or the choice of language(s), like Latin for medieval manuscripts intended for use in the Roman liturgy. The choice of images that may be included in manuscripts in certain contexts, for example the Evangelist portraits in Gospel Books, can also be counted among the content-related patterns. In liturgical Psalter manuscripts, the canticles usually follow the 150 psalms, as does a litany. Additionally, the psalms are preceded by a calendar. The other category of patterns includes the choice of writing material (such as parchment and paper), (painting) techniques used or the visual organisation. Patterns can indicate how to use a manuscript: the direct juxtaposition of two translations of a text in two columns invites comparison, while in other manuscripts a reference system is formed by placing individual components (such as a section of text) in an overall context (like a chapter). In the latter case, direct access to specific sections may be enabled by initials or marginal notes, among other things, so that the manuscript can be used for teaching purposes, for example.15

• 16 Although Kwakkel does not speak specifically of frames and patterns, he points out that the materia (...)

8It has been rightly emphasised that many patterns are highly conventionalised so that “neither the producer(s) nor the user(s) of the manuscript will give them much thought” (Wimmer et al. 2015, 9).16 It is a well-established convention that a Latin text is read from left to right and from top to bottom. This is reflected in text production in that writing usually intuitively begins at the top left of a page. The focus of this paper will be on more unusual patterns that can be associated with the Divine Office and the potential use of manuscripts in this context. While it was pointed out that patterns are structures and conventions that organise the manuscript and make it accessible for its user(s), there has been insufficient discussion of what exactly is to be understood as a pattern, especially in terms of visual organisation. It can be argued that, for example, each initial is in itself a pattern by which the manuscript is structured and the beginning of a section is emphasised. In the case of Psalter manuscripts, however, in which individual psalms are each marked by an initial, it makes more sense to interpret a set of units associated with a specific frame as one pattern. In the following, a set of prominent initials that are important for the Divine Office will be discussed as an example. These initials are collectively to be understood as one (admittedly complex) pattern.

• 17 There are other ways of dividing the Book of Psalms and grouping Psalms than the three mentioned he (...)

9It is well known that in most manuscripts, the Book of Psalms was divided into groups by means of larger initials. In the twelfth, thirteenth and fourteenth centuries, from which most of the manuscripts discussed below date, three types of subdivision are particularly common: a formal division into three parts with 50 psalms each, a liturgical division into eight parts corresponding to the basic distribution of psalms over the week in the Divine Office, and a combination of both resulting in a tenfold division (not elevenfold, since Psalm 1 marks the beginning of the first group in both systems).17 The eightfold division is closely tied to the Divine Office: Psalms 1 to 108 (with several exceptions) were recited at Matins over the course of one week; they are therefore subdivided into seven groups, one for each day (e.g. Psalms 1–25 or Psalms 26–37). The eighth group consists of all the psalms sung at Vespers during the week (without further subdivision). The first seven groups begin with Psalms 1, 26, 38, 52, 68, 80 and 97, and the eighth group starts with Psalm 109.

• 18 For a detailed discussion on how the Book of Psalms was divided in the Luttrell Psalter, see Becker (...)
• 19 In many manuscripts, including the Luttrell Psalter, the initial to Psalm 1 has a special position (...)

10As I have argued elsewhere, the tenfold division is not always a pattern in its own right. It can also be a combination of the two other patterns, the threefold and eightfold division, which remain clearly distinguishable in the manuscript. This is the case for the fourteenth-century Luttrell Psalter (London, British Library, Add MS 42130).18 In the Luttrell Psalter, regular psalm initials extend over two lines, while the initials to the ten divisions usually occupy four lines in height. There are three exceptions, i.e. the Beatus initial to Psalm 1 as well as the initials to Psalms 51 and 101.19 The latter two extend only over three lines instead of four (see figs 1 and 2). Therefore, the Luttrell Psalter does not highlight the beginnings of the ten sections uniformly. Rather, by looking at the initials, one can distinguish, between the two different patterns, the formal tripartite division on the one hand and the liturgical division into eight groups on the other.

• 20 As I have mentioned elsewhere with regard to the potential usability of Psalter manuscripts in gene (...)

11In order to decipher these patterns and connect them to an abstract frame, the reader of a Psalter manuscript would have to be familiar both with the manuscript as such and with the ritual practice, the Divine Office, as a potential context of use. With regard to the first point, a relative difference in size of three to four lines, for example, can easily be overlooked as a marker of different patterns if the user has not studied the manuscript thoroughly. It is even possible that the reader thinks the slightly different sizes are a mistake on the part of the producers. At the same time, the reader must have at least a basic knowledge of various cognitive frames, such as the ritual practices in which the manuscript could be used. Without such knowledge, different patterns may be identified, but the reason for their inclusion and which frame they evoke would remain hidden. For example, if the psalms are divided into eight groups corresponding to the general distribution of the psalms in the Divine Office, the large initials structure the codex and make the beginning of each of the eight groups easily accessible. Nevertheless, the reader does not necessarily understand why exactly the Book of Psalms is divided in this way if they cannot link it to the Divine Office. Why are there eight groups? Why are there twenty-five psalms in the first group and only twelve in the second? It is unclear whether and how Psalter manuscripts were actually used and read during the Middle Ages, especially those that include additional textual material that was needed for the Divine Office.20 The patterns that can be traced in the manuscripts can provide a basis for reconstructing these abstract frames of use, although it should be noted that this is not proof that the manuscripts were actually used in this way.

• 21 For a detailed record of the manuscript, see Plotzek 1987, 66–67 (Cat. no. 1) and Sauer 1994, 60–61 (...)
• 22 These are the initials to Psalms 114 (fol. 115v), 121 (fol. 127v), 126 (fol. 129r), 131 (fol. 131r) (...)

12From here on, I will discuss different examples of both content-related patterns and patterns related to visual organisation, all of which suggest that the manuscripts could be used in the Divine Office, or at least were closely related to this ritual practice. The eightfold division that allows easy access to the first psalms sung at Matins has already been mentioned above. The eighth group, which comprises all the psalms sung at Vespers in the course of the week (beginning with Psalm 109), often has no further distinction that would allow the user of the manuscript to access, for example, the first psalm sung at Vespers on Monday (Psalm 114) directly. A rare counterexample, in which a subdivision of the eighth group can be observed, is the manuscript Cod. Don. 180 (Stuttgart, Württembergische Landesbibliothek).21 Within the group of vesper psalms, the opening psalms of the seven days of the week are marked by initials that are painted over in silver (see fig. 3).22 This visibly distinguishes them from the otherwise red psalm initials.

• 23 For a detailed record of the manuscript as well as the digitised manuscript, see the British Librar (...)
• 24 One of several other examples for this is the Luttrell Psalter (fols 257r–259v). However, here the (...)
• 25 In the Liber officialis (or De ecclesiasticis officiis) of Amalar of Metz (c. 775–c. 850), the fact (...)
• 26 In fact, the canticle is numbered as Psalm 148, not 149 as one would expect, as a consequence of an (...)

13Not only can individual psalms within a section be highlighted, but individual psalms have also been joined together. In the manuscripts considered here, however, this could only be observed for Psalms 148–150 (or in one case Psalms 149 and 150). “Joined together” in this case means that only one psalm initial has been painted, i. e. the initial for Psalm 148, while Psalms 149 and 150 follow as if they were additional verses of Psalm 148. This can be observed, for example, in the Gorleston Psalter (London, British Library, Add MS 49622):23 Towards the end of fol. 188v, Psalm 148 begins with a two-line initial (see fig. 4). The psalm ends on fol. 189v after about two-thirds of the page and Psalm 149 follows directly (see fig. 5). Unlike regular psalms, however, it is not decorated with a two-line initial, but with a one-line initial. It can only be distinguished from verse initials because it is not purely ornamental but has a face added to it. The same is true for the beginning of Psalm 150 (fol. 190r).24 In both the secular and monastic cursus, Psalms 148–150 were part of the daily Lauds, where they were recited continuously and without interruption. In this case, this means that no antiphons or the doxology were sung between the individual psalms, as was otherwise customary. Instead, the three psalms were treated as one (e.g. Hughes 2004, 231).25 The combination of Psalms 148–150 into what appears to be a single psalm presented a later user of a fourteenth-century manuscript from the Bayerische Staatsbibliothek in Munich (Clm 16137) with a challenge: when numbering all the psalms in the margins so that they were easier to find, the later user counted the first canticle – which follows the Book of Psalms – as another psalm, since Psalms 148–150 (fols 136v–137v) were perceived as a single psalm.26 This already shows that patterns can also be misinterpreted and I will address this in more detail below. The fact that these three psalms are combined into a single psalm in several manuscripts shows the influence of ritual practice, i.e. the Divine Office, on the manuscripts themselves. The textual sovereignty of the individual psalms is broken down in favour of the manuscripts’ (potential or intended) use during the Divine Office.

• 27 See also Harper 1991, 300; and Hughes 2004, 24.
• 28 I have discussed this in more detail elsewhere with further examples (Becker 2024, 293–94).
• 29 A record of the manuscript is available online at the Boston Public Library (Boston Public Library, (...)

14At least as far as I am aware, two interesting content-related patterns have often been overlooked and not discussed in detail. Both patterns are rare, but not singular. As already indicated above, additional elements such as canticles and scriptural lessons were recited in the Divine Office. One key element was the so-called lesser doxology (Gloria patri), “a form of praise to God,” “recited at the end of most psalms, canticles, and hymns, and near the beginning of most Offices” (Harper 1991, 297).27 It was occasionally inserted in Psalter manuscripts after the Book of Psalms. In the Luttrell Psalter, for example, it follows the canticles (fol. 283r), without being specifically marked as the lesser doxology. In some manuscripts, however, the doxology has been inserted more than once, directly after individual psalms.28 The doxology was not inserted after each and every psalm, as one might expect given its use in the Divine Office. On the contrary, the number of times the doxology was inserted and its placement in the manuscript seem arbitrary. A Psalter manuscript from the Boston Public Library (MS f Med.84) made around 1250 contains the doxology three times:29 It precedes the two large initials of Psalms 52 (fol. 62r) and 97 (fol. 103v) that mark the beginning of a new division as well as the regular psalm initial of Psalm 84 (fol. 92r). In places where the next initial would no longer have fit on the page because of its size and where it was therefore moved to the top of the next page, the doxology is used to fill the lines that would otherwise have remained empty. On fol. 61v the doxology takes up two lines, before the more than half-page initial to Psalm 52 follows on the next leaf (see fig. 6). From what we can see, the doxology in this example has been inserted primarily for aesthetic reasons. However, to fill the page and avoid blank lines, the illuminators or scribes could just as well have used line fillers or other texts. But instead, they have opted for the Gloria patri. They thus created a content-related pattern that was directly linked to the recitation of the psalms in the Divine Office, and its inclusion can be understood as a reference to this cognitive frame. This connection would not have escaped a reader who was familiar with the Divine Office and the function of the doxology within it. Even though the doxology seems to appear in the manuscript for primarily aesthetic reasons, it is nonetheless inextricably linked to the cognitive frame.

• 30 On the literary structure of Psalm 118, see Callaham 2009.
• 31 For the general distribution of the psalms in the Divine Office, see, as already mentioned above, t (...)
• 32 A more detailed look at these initials would be worthwhile. Some of them are nevertheless distingui (...)

15Another example is Psalm 118, with 176 verses the longest of all. In the Bible, it has been divided into twenty-two sections, each of which contains eight verses and is marked with a letter of the Hebrew alphabet.30 In both the secular and the monastic cursus, one or more of those sections of Psalm 118 were recited at different hours. For instance, Psalm 118,33–80 (the fifth to tenth sections) were recited daily at Terce in the secular cursus.31 The biblical division into twenty-two sections is usually adopted in medieval Latin Psalter manuscripts and manifests itself visually in additional psalm initials. That is, each of the sections (or sometimes every other) has its own initial, corresponding to the psalm initials in size, although it is not a new psalm that begins, but rather a new section of Psalm 118.32

• 33 For the complete Latin text as well as a German translation, see Drecoll 2007, 54–56.
• 34 On the manuscript, see Kuder 1996 (Cat. no. 31). The digitised manuscript is available online (Würt (...)

16The division of Psalm 118 is of particular interest here because in two Psalter manuscripts produced in Germany, an additional text was inserted between two of the sections. The text in question is the Athanasian Creed (Quicumque vult, also sometimes Quicunque vult).33 In a thirteenth-century manuscript (Stuttgart, Württembergische Landesbibliothek, HB II 25), which was presumably produced in Franconia, the Athanasian Creed was added in between the fourth and fifth sections of Psalm 118, i.e. between verses 32 and 33 (fols 92r–93v).34 The Athanasian Creed has an initial of the same size as the psalm initials, and thus cannot be distinguished visually from either the actual psalms or the sections of Psalm 118 (see fig. 7). It was recited at Prime on Sunday, where it followed Psalm 118,32 in both the monastic and secular cursus (Harper 1991, 100, 312; Hughes 2004, 38, 76). Thus, the Athanasian Creed was not only added in the middle of a psalm instead of after the Book of Psalms, together with the canticles and other devotional material. But above all, it was added in exactly the place where it would have been recited if the manuscript had been used in the Divine Office. I have only observed this so far in the two manuscripts mentioned, but it would certainly be worthwhile to re-examine other manuscripts with this in mind. As already mentioned, patterns can only enable or facilitate a particular use of the manuscript if the reader of the manuscript is able to perceive them. The text is a content-related pattern and its placement, in this case, is an additional link to the Divine Office. Since it was not combined with visual aids like a rubricated title, it might have been too well hidden even for a contemporary reader.

• 35 In the secular cursus, for instance, Psalms 4 and 5 were not recited during Matins on Sunday, but r (...)

17Although the patterns in the manuscripts do invite certain interpretative frames and enable potential readers to use them, there are noticeable difficulties with regard to the usability of the manuscripts. Most Psalter manuscripts contain structuring aids such as different sized initials for the divisions, psalms and single verses. While the division into eight parts with the help of large initials allows direct access to the groups of psalms recited at Matins on the seven days of the week, this is only a simplification of the actual use of the psalms in the Divine Office. The most obvious problem here is the fact that several psalms were recited at different hours and not in numerical order.35 Sometimes they were even recited more than once a week. I have discussed this in detail in another article (Becker 2024, 291-93), thus a few brief remarks will suffice here. Psalm 94, for example, was recited daily at the beginning of Matins as the invitatory psalm (the psalm sung as an “invite” to prayer) rather than being recited after Psalm 93 and before Psalm 95 during Matins on Friday (e.g. Harper 1991, 70; Hughes 2004, 51; Billett 2014, 15). What becomes clear is that, while the pattern of the liturgical eightfold division to a certain extent enables and even suggests a specific use of the manuscript, individual psalms within the eight groups that were needed in the Divine Office were usually not highlighted. It was and is therefore difficult to access them directly. One can only agree with Duffy in that Psalter manuscripts “presented unique problems of navigation for the lay (and indeed clerical) user” and that “the layout of medieval Psalters […] militated against easy reference to individual Psalms other than those which headed the traditional eightfold or tenfold division of the Psalter” (Duffy 2008, 93).

• 36 A short general overview on Breviaries is found in Palazzo (1998, 123–25). On Books of Hours in gen (...)
• 37 Rudy distinguishes between two categories of additions: additions that require rebinding and those (...)
• 38 On the manuscript, see the detailed bilingual commentary (English and German) by Sandler (1999). Fi (...)
• 39 For a list of all the German headings, see Sandler 1999, 110–11.

18The manuscript architecture of Psalter manuscripts and the visual organisation of the individual pages are highly complex. They require a user who is not only familiar with the manuscript in question, but also with the ritual practice. Even then, not all aspects and details that would be necessary for their use are taken into account. At the time when most of the Psalter manuscripts discussed here were produced, two types of books already existed that were much easier to use: the Breviary in which the elements of the Divine Office were grouped together and the Book of Hours for use of the laity.36 The development and increasing popularity of these two types of books, however, did not initially result in Psalter manuscripts no longer being produced. They were still produced in large numbers in the thirteenth and fourteenth centuries, but they could be reworked by one or more later owners to personalise them, to include new texts and/or to commemorate other feasts and saints etc.37 Some of the changes can be attributed to a shift in use. This is the case, for example, in the Ramsey Psalter (Stiftsbibliothek St. Paul im Lavanttal, Codex 58/1).38 German rubrics were added throughout the manuscript in the late fifteenth century. They “mark the beginning of each group of psalms recited at Matins and Vespers on the various days of the week, as well as those recited daily at Sext or None” (Sandler 1999, 110).39 Psalm 68, for example, was recited in the secular cursus as the first psalm in Matins on Thursday and this is exactly what is written in the margin next to Psalm 68 in the Ramsey Psalter (fol. 76r): “Nocturn am Dornstag.” While the manuscript was originally produced for the Ramsey Abbey (Cambridgeshire, England), a Benedictine abbey, the German rubrics follow the secular rather than the monastic cursus. Thus, they “reveal a shift in the use of the Ramsey psalter” (Sandler 1999, 111).

• 40 In some manuscripts, this led to errors in numbering that had to be corrected later. For example, i (...)
• 41 In her catalogue of the Latin liturgical psalters in the Bodleian Library, Solopova repeatedly draw (...)

19Other additions seem to address a perceived lack of handiness – at least from our modern point of view – by numbering the individual psalms at some point after production. In the aforementioned manuscript MS f Med.84, numbers were added next to every psalm (see fig. 6). According to the library’s catalogue, they date back to the fifteenth century (Boston Public Library, “MS f Med.84”). These numbers form an additional new pattern that facilitates access to individual psalms, thereby referring either to a different (or at least altered) frame of use or changed requirements on the part of the reader. Numbers generally make it easier to navigate through the book, especially with the “I” initials, which tend to merge with the marginal decoration in many manuscripts and are easily overlooked.40 A reader would therefore no longer have to read the text and know it by heart in order to know which psalm they are reading. What we do not know is to what extent external aids such as bookmarks were used by early readers to make individual sections or psalms accessible. In several manuscripts, (remains of) glued-in leather or parchment strips marking individual leaves have survived.41 Loose strips of parchment and the like, which served as bookmarks, can indeed still be found in the manuscripts, though they have certainly been lost in many cases.

• 42 The Roman numerals for Psalm 118, 153–160 (fol. 131v) are missing, although the section before is n (...)

20While the later numbering helps the reader navigate through the manuscript, MS f Med.84 is a perfect example of how patterns can be misinterpreted. Psalm 118 is divided into twenty-two sections in MS f Med.84, each with a psalm initial. For Psalm 118 (fol. 124r), the corresponding number in Roman numerals has been added in the margin (see fig. 8). On the same page, the second section (verses 9–16) begins with a large initial “I,” next to which are the Roman numerals for 119. The other sections of Psalm 118 are likewise numbered until the numbering breaks off with the number 138 at the penultimate section of the psalm (verses 161–168).42 The person who did the numbering did recognise the pattern of four-line initials for the beginning of a new regular psalm. However, because the section initials to Psalm 118 form a pattern on their own that is identical on a visual level, the person failed to recognised that these are two different patterns. Whoever added the numbers apparently did not pay attention to the text but rather just added a number next to every initial, assuming that a new psalm begins with each four-line initial. A later user of the manuscript thus failed to recognise or misunderstood an original pattern, just as in the previously mentioned example with the misnumbering of Psalms 148–150.

• 43 For further examples, see also Becker 2024.

22A close intertwining of manuscript and Divine Office is evident in the manuscripts examined in this paper.43 The cognitive frame, in this case the Divine Office, has consciously or subconsciously found its way into the manuscripts and is manifested visually in the form of patterns, expressed in various forms of book illumination. It shows that those who made the manuscripts must have had an extensive knowledge of the cognitive frame in order to include diverse patterns. Although it is possible that they merely copied another manuscript without understanding the meaning of each pattern, the varied and individual solutions outside the common division into eight parts indicate that the producers were not only familiar with the Divine Office, but also consciously incorporated it into the manuscripts. To use the concept of frame analysis in manuscript studies can lead to a better understanding of the motivations behind certain design decisions in manuscripts and their purposes. It can also help discover more and more subtleties of the visual organisation and see meaning in what seems to be arbitrary at first. This approach also implies some advice of caution, as there might be even more patterns found in manuscripts and associated (cultural) frames of which we are currently unaware.