Eighth-century Insular Gospels (NLR, Lat. F.v.I.8): Codicological and palaeographical aspects

Olga Bleskina, The National Library of Russia

The text of this article is also available in Russian.

Images reproduced with the kind permission of the National Library of Russia.

1. Introduction

The Insular Gospels of the 8th century (NLR, Lat. F.v.I.8) from P. P. Dubrovsky’s collection is one of the most famous western manuscripts in the collection of the National Library of Russia. Ornamented with splendid decorated initials in the insular animal style, this manuscript is one of the best examples of early book illumination. It is dated back to the second half – late 8th century and considered to have been produced in England under the influence of the Irish book art.

The investigation of the Insular Gospels began in the 18th century and still continues. [1] Among many scholars who have studied the manuscript are N. R. Ker, E. A. Lowe, O. A. Dobiaš-Rozdestvenskaja, T. D. Kendrick, R. Kockelkorn, and others. The researchers were equally attracted by the decoration of the manuscript and its script. It may seem that everything has already been described; however, the codicological analysis of the manuscript still requires more precise definitions: how many people were involved in the production of the codex and how was the whole work carried out?

2. The material of the script

The Insular Gospels were written on thick parchment of such fine treatment that the difference between the flesh and hair sides of the folios is almost imperceptible. Special selection of folios – which was the practice of later times – was not necessary for high-quality parchment: bifolios could be folded into quires without taking sides into account, so that in the quires of this codex, flesh and hair sides appear next to each other.

3. The structure of the codex

The codex consists of 215 folios (345 × 245 mm), but the original folios used to be larger and were only cut from the top and sides during the second binding. The final folio 214 containing prayers was added to the codex in the 11th century. The manuscript has 29 quires including 18 quaternions, one quinion, and three trinions. [2] Ten quires also contain folios apart from bifolios. This may suggest that the structure of the codex was being changed in the process of work (see Table 1).

Table 1. The structure of the codex.

Number of quire	Numbers of folios	Number of folios in quires	Composition of quire	Notes
1	1-6	6 (binion + 2 ff.)		earlier – trinion; a folio is cut from folio 2, folio 6 is inserted
2	7-11	5 (binion + 1 f.)		an empty folio is cut from folio 7 (last folio with contents ending)
3	12-17	6 (trinion)
4	18-27	10 (quinion)
5 6 7 8 9 10	28-35 36-43 44-51 52-57, 57bis, 58 59-66 67-74	8 (quaternion)		folio 57bis is missed in old foliation
11	75-77	3		separate folios (inserted)
12	78-85	8 (trinion + 2 ff.)		folios 80 and 83 are inserted
13 14 15	86-93 94-101 102-109	8 (quaternion)
16	110-118	9 (quaternion + 1 f.)		folio 112 is inserted (the beginning of the text “Argumentum”), is written by another pen
17	119-126	8 (quaternion)
18	127-133	7 (trinion + 1 f.)		a single folio 129 is inserted (instead of incorrect?)
19 20 21 22	134-141 142-149 150-157 158-165	8 (quaternion)
23	166-171	6 (binion + 2 ff.)		earlier quaternion; folios with text are cut from folios 167 and 168 (text fragments are visible), folio 171 – another handwriting
24	172-176	5 (binion + 1 f.)		single folio 175 is inserted, written with another handwriting
25 26 27	177-184 185-192 193-200	8 (quaternion)
28	201-208	8 (binion + 4 ff.)		folios 202, 204, 205, 207 – single folio
29	209-213	5 (bifolio + 3 ff.)		folios 209, 210, 211 – single folio

To be able to reconstruct the process of codex production and to define the number of people who took part in the work, I have to analyse the types of ruling and supplementary elements, such as signatures and marginal marks. I shall first consider the ruling – the preliminary stage of work.

E. A. Lowe has pointed out that the ruling was done with a blunt point after the folios had been folded into quires. Each folio is ruled separately up to the preface to Mark (f. 75), and then the ruling is only seen on every second or third folio. To the left of each column there are double bounding lines, single lines to the right. [3] So, according to Lowe, there is only one type of ruling. E. L. Privalova has suggested that there was one more: two horizontal lines on the right and another two on the left. [4] In fact, three types of ruling can be distinguished in the manuscript (see Illustration 1).

The first type is characterised by: (1) line and marginal prickings are knife-cut, (2) prickings are located in the margins and badly distinguishable, (3) the ruling is made exactly to the prickings, (4) there are double vertical lines for initials in each column on the right and left, (5) the ruling lines go into the margins as far as the sheet edges. The ruling was made after the sheets had been folded into quires. Several sheets were ruled on one side at a time.

The second type: (1) pinpricks are made by a penknife or a quill, cut prickings are only found in quires 7-9, (2) marginal prickings are not distinguishable in all quires, (3) double vertical lines for initials are ruled only on the left side of each column, (4) ruled lines go beyond the pricks, but do not reach the edges. The ruling was made after the sheets had been folded into quires. All the sheets were pricked at the same time, and then each sheet was ruled separately on one side. The ruling of quires 7-9 took two stages: at first, knife-cut pricks for the margins were made and then pinpricks for the lines.

The third type: (1) line and marginal pinpricks are made by a quill, (2) the prick points are only distinguishable on the farthest bounding lines of the margins, (3) there are double vertical lines for initials in each column on the right and left, (4) vertical ruled lines hardly go beyond the margins, (5) the line ruling is made exactly to the prickings and only within columns. The ruling was made after the sheets had been folded into quires. All the sheets were pricked at the same time, and then every two sheets were ruled on one side.

This suggests that the quires were prepared by three persons. All of them ruled the sheets after they had been folded into quires. Each person applied his own way of ruling and used his own instruments. The first of them made knife-cut prickings, the second one used a knife and a quill, the third just a quill. Some of the sheets contain double and triple prickings and ruling corrections. This is more frequently observed in places where one script supersedes the other and on separate sheets that were added to the codex at a later stage.

4. Supplementary elements in the structure of the codex

Most of the scholars who have studied the manuscript mention that it contains signatures and that the folio numbering starts only from quire 12 (the Gospel of Mark). In fact, the signatures in the form of Roman figures from “.I.” to “.IIII.” and from “.UI.” to “.XUI.” are stated on the final verso of each quire (12-15, 17-23, 25-28) by a scribe, who prepared the folios for copying. These quires contain the texts of the Gospels of Mark, Luke, and John, excluding the prologues. The quires had been numbered before the text was copied. The signatures are only missing from quires 16 (signature “.U.”) and 24, which can be easily explained if we take a look at Table 2. These quires are not complete: they were formed later when the prologues were added to them, and were partly copied by a different scribe, who was not using signatures.

The codicologists have pointed out that the upper margins of folios in the Gospel of Matthew contain small crosses and letter combinations XB, +BXB, and +XB. D. Wright noticed that these letters were located on the first folio of the quires 4-10. [5] I suggest that these combinations are also used as signatures. The letter signatures are found in the beginnings of five out of seven quires, and the crosses carry on the marking. Such marking is not used on the first illuminated folios of quires 4 and 8. This numbering system is only characteristic of the Gospel of Matthew. The incomplete quires and the inserted prologue sheets are also excluded from the numbering. [6]

5. The script

The Insular Gospels are written in a large calligraphic script, called by the scholars majuscule [7] or half-uncial [8] with a geographic specification – Insular or Anglo-Saxon. [9] As the script of the Gospels contains ascenders and descenders and even minuscule forms, it seems more appropriate to term it as Insular half-uncial script. The Insular half-uncial was spread in Irish and English scriptoria in 6th–9th centuries. This elegant calligraphic script is characterised by small ascenders and descenders not going beyond the x-height, rounded shapes, square proportions, lack of slant, triangular initial letter elements, and the use of uncial and minuscule forms for some letters. [10]

The Insular Gospels from the NLR collection meet all these requirements. The script is straight without slant, the letters are rounded, the ascenders and descenders are straight, their height being almost two times smaller than the x-height. The number of ascending/descending elements is minimal. The serifs of most of the letters are triangular. The texts of all Gospels, excluding the prologues, are written per cola et commata.

At first sight, the script looks uniform, but a closer look at it shows a number of specific features in different parts of the codex in terms of calligraphy, graphics, and even size. All the scholars who studied the manuscript said that the Insular Gospels had been written by different scribes, but they were not able to tell how many people had been involved in the production of the codex. [11] It was E. L. Privalova who suggested that it had been a work of three artists: the first artist is responsible for the Gospel of Matthew, the second worked on Mark, Luke, and John, while the third – according to Privalova, he was less qualified than the other two – produced the illuminations of the prologues. [12] They could be the same people who copied the text; however, Privalova was not able to distinguish between several scribes.

The palaeographic analysis shows that the manuscript was indeed copied by three scribes (see Illustration 2). Let me consider them in more detail.

The first hand (ff. 1-11v, 75-77v, 113-118, 171-176v) is smaller and less calligraphic than the other two. It is the hand of all the prefaces and prologues, except the prologue to the Gospel of Luke. It is characterised by the most extensive use of minuscule letters, by smaller proportions and rounded shapes. [13] The hand of the first scribe is lighter: he was using a finer pen and almost black ink (see Illustration 2/1 and Illustration 3, right-hand side).

Here are the main features of this script. Letter shapes: d is of just one shape with a straight vertical ascender; e is closed in a shape like the Greek θ, with an extended middle line at the end of words; g has a curved descender without a loop; n has two shapes: minuscule and, very seldom, majuscule, typical for early manuscripts, found both at the beginning and in the middle of words; r is shaped as a minuscule n and very seldom as majuscule R; s has two shapes: an ordinary short s and a final ſ with ascender, characteristic of the Insular minuscule; u and v are shaped as minuscule u. (See Figure 1.)

The first scribe employs ligatures occasionally and punctuation marks – punctus elevatus (a long pause) – at the end of some sentences. The word division is mostly regular, with abbreviations of several types: special symbols, contractions, Nomina Sacra, and occasional suspensions.

The Gospel of Matthew (ff. 18v-74v), the first Gospel in the codex, is written by the second hand, the most calligraphic and proportionate (see Illustration 2/2). It is characterized by rounded uncial letters with elegant curved verticals. The script is a bit heavy, with square-proportioned letters, but because of the rounded shapes it does not look ponderous. The second scribe used solid dark-brown ink.

The script of Matthew has the following characteristics: elegant b and l with rounded shapes and small ascenders with triangular serifs; d has two shapes: one with a straight vertical ascender and an open oval, and the other with a slanting horizontal ascender and a closed oval; e is open at the bottom with an extended middle line; g has a closed loop; r has only majuscule shapes; s is always short; t has an insular rounded shape and seldom a majuscule T shape at the end of lines; u is shaped as u and sometimes as v. (See Figure 2.)

The second script does not employ punctuation, except quotation marks for citations from the Old Testament, and only very seldom employs abbreviations. The word division is regular, but many combinations of nouns, pronouns, or adverbs with prepositions are spelt as one word.

The third copyist (Figures 3–5) is responsible for almost two-thirds of the codex: the Gospels of Mark, Luke, and John (ff. 78v-112v, 119-170v, 177v-213v). This script looks as if it belonged to several hands and is difficult to identify (see Illustration 2/3 and Illustration 3, left-hand side). Very big, close to majuscule at the beginning of Mark, it gradually changes proportions, sometimes getting smaller and looser towards the end of a quire, sometimes being more calligraphic again. It is the heaviest hand in the manuscript: straight square-proportioned letters are written with a thick pen with a wide cut, pen and ink changing several times within one hundred or so folios.

Here are its main features: a has three shapes; d has two shapes: one with a straight ascender, and the other with a slanting ascender; e has three shapes: ordinary, uncial with an ascender, and a closed one in a shape like the Greek θ at the end of words; g has a closed loop; n has minuscule and majuscule shapes; r has three shapes: an ordinary straight shape, minuscule n shape, and majuscule R shape; two kinds of s: an ordinary short s and a minuscule ſ with ascender; t has insular or majuscule T shape at the end of lines; u is either ordinary or majuscule v.

The third scribe uses ligatures of majuscule and mixed letter shapes, and two types of abbreviations. The word division is mostly regular. The only punctuation employed is quotation marks – colour dots – for citations. His specific feature is the use of phrases written in a different script – Insular minuscule – at the end of lines. [14]

The palaeographic analysis of the Gospels of Mark, Luke, and John shows that in spite of the visual differences, they belong to the same scribe.

6. Some peculiarities of the decoration

The decoration of the Insular Gospels has been studied by many scholars, so here I shall only concentrate on those details of ornamentation which can help us to reconstruct how the codex was produced and who were the people involved in this production. Art historians think that the codex was illuminated by three artists with varied degrees of qualification: the first artists produced the initial page of Matthew (f. 18) and the arcades (ff. 12-17), the second one made the initial pages of Mark and Luke (ff. 78, 119), while the third is responsible for the ornamented page in John (f. 177). [15] The work of each artist is distinguished by his own technical style, level of qualification, and colour scale.

I suggest that the decoration of the Insular Gospels (initial pages with frames, arcades, and ornamented initials) was done by two people – the first and the third scribes who copied the text. The first copyist produced not only the arcades with canon tables and initial pages of Matthew, but also the decorations of the first three quires and the first folio in quire 4 (ff. 1-18). He also illuminated the text in quire 11, part of quires 16 and 24, and the first folio of quire 25, that is the beginning of John (ff. 75-77v, 113-118v, 172-177) (see Illustration 4). [16]

According to many scholars, this artist is the most professional. He is very good at producing elegant filigree initials and decorated frames, and his choice of colours is done with taste. His colours are both bright and harmonious: chrome, minium, other hues of red, light-brown, and dark-blue. The first illuminator uses red and black dots to frame initials, and triangular dots to fill the background in decorated frames. [17]

The decorations of Mark and Luke – initial pages, middle and small initials, and the initials in John, excluding the first folio – were, in my opinion, made by a different hand (see Illustration 5). His style is simpler and technically less elegant. He uses bright and variegated colours, e.g. green, which are not found in the palette of the other artist. The colour scale of the initials corresponds to that of the initial pages in these Gospels. The ornamented frames in ff. 78 and 119 are decorated with a delicate pressed pattern containing dots against the white background. The same pattern is used in borders with white rectangles and no colour enclosing, which produces an unfinished effect. Small initials are decorated with traditional dots, and some of them have additional colour spots shaped as dotted suns.

My observations are summarised in Table 2. This data may allow us to conclude that three scribes took part in the production of the Insular Gospels. They worked on their quires in stages – from ruling to decoration. All types of work each of the copyists made himself: he prepared the quires for writing, made prickings, ruled margins and lines, copied the text and decorated it.

The first scribe prepared the quires for the copying, copied and decorated all the prefaces, prologues, and tables of contents (except the prologue to Luke), produced the arcades with canon tables and initial pages of the Gospels of Matthew and John – quires 1-3, 4 (f. 1), quires 11, 16 (f. 6), 23 (f. 1), 24, 25 (f. 1), or ff. 1-18, 75-77v, 113-118v, 171-177. His typical features are special slash knife-cut prickings and ruling with double bounding lines for initials to the left and to the right of each column, relatively small script with predominance of minuscule letters, a special style of decorated initials with blue and silver colours – a feature which is not attested in the ornamentation elsewhere in the codex.

The second scribe ruled, copied and coloured small initials in the Gospel of Matthew: quires 4-10 (ff. 18v-74v). He employs a different type of prickings (round prickings made with a pointed instrument), double bounding lines for initials to the left of each column, a special type of signatures, calligraphic rounded script, and just two colours in small initials (chrome and red ochre).

The third scribe did the largest portion of work – he prepared, copied, and illuminated the Gospels of Mark, Luke and John (except the initial page of John, which was ornamented by the first scribe). These are quires 12-15, 16 (f. 3), 17-22, 23 (f. 5), 25-29 or ff. 78-112v, 119-170v, and 177v-213v. The peculiarities of his style are the following: he does not use marginal prickings, his ruling is casual, line ruling is only used in columns, the quires are signed with numerals (but not evenly), the script is big and heavy, and the decoration is stylistically different from the other scribes: it uses bright and variegated colours (e.g. green and lilac).

The copying started from the Gospel texts, which were written by the second and third scribe (in my classification). Their tasks divided, the scribes were working at them on their own. As the repertory of prologues was not fixed by that time, prologues and prefaces were only included into a codex at the final stage. Portions of ready texts had to be replaced then, e.g. the final verses of Luke in the prologue to Mark, following f. 170. This can also be proved by a number of incomplete quires, added and cut folios with fragments of cut-off texts. The prologues were copied by the first scribe (in my classification). As already mentioned, he produced the initial pages of Matthew and John, and canon tables. It appears that this scribe was the most experienced, the one who formed the codex after its parts had been copied, simultaneously playing the role of its corrector. However, this problem, as well as many others in the study of this unique manuscript, remains to be solved.

English translation by Olga Timofeeva

Notes

[1] For the first time this codex is mentioned in the famous work of B. Montfaucon, “35. 1057. Evangelia literis Saxonicis” (Montfaucon B. de, Bibliotheca bibliothecarum manuscriptorum nova… Parisiis, 1739. T.II. Col.1142). The most important researches for this theme: Tassin & Toustain (1757: 226-230, 380-382); Lindsay (1915: 58, 182, 249); Zimmermann (1916: 304-305), Wright (1961: 122-123); Privalova (1962: 201-211); Lowe (1966: 7, № 1605); Dobiaš-Roždestvenskaja and Bakhtine (1991: 58-61, № 26); Kilpiö and Kahlas-Tarkka (2001: 41-44); Bleskina (2007: 303-327). For more complete bibliography, see Elagina et al. (2001).

[2] The correct composition of the quires was only made by D. Wright. However, he considered that the 13th quire should be a quinion and consist of 10 folios, but it is not confirmed textologically. See Wright (1961: 122-123).

[3] Lowe (1966: 7, № 1605).

[4] Privalova (1962: 210).

[5] Wright (1961: 122-123).

[6] The composition of the prologues to the Gospels was not fixed and was often changed in the 8th century. Probably the text was replaced in the course of the work on the Insular Gospels. Fragments of the cut texts are visible on folios 167 and 168.

[7] Lowe (1966: 7, № 1605).

[8] Dobiaš-Roždestvenskaja and Bakhtine (1991: 58).

[9] Anglo-Saxon half-uncial: Zimmermann (1916: 304); Anglo-Saxon majuscule: Wright (1961: 123).

[10] On Insular half-uncial see Bischoff (1979: 108-115), Brown and Lovett (1999: 51-58).

[11] Zimmermann (1916: 304); Lowe (1966: 7, № 1605); Dobiaš-Roždestvenskaja and Bakhtine (1991: 58); and others.

[12] Privalova (1962: 210-211).

[13] M. Braun names such scripts as Insular hybrid minuscule. See Brown & Lovett (1999: 53-54).

[14] This is a characteristic feature of the Irish script.

[15] Zimmermann (1916: 304-305). E. L. Privalova considers that the first artist performed the arcades with canon tables, the initial page of the Gospel of Matthew, the framed page with the initial of John; the second artist, the decoration of the Gospels of Mark and Luke; the third one, the text and decoration of John (Privalova 1962: 210-211).

[16] Folio 171 is a later addition and is not decorated.

[17] A specific feature of the Irish manuscripts decoration.

References

Bleskina, O. N. 2007. "Ostrovnoye evangeliye VIII veka (RNB, Lat. F.v.I.8): Kodikologicheskiy i paleograficheskiy aspekty [Eighth-century insular gospel (NLR, Lat. F.v.I.8): codicological and palaeographical aspects]". Monfokon: Issledovaniya po paleografii, kodikologii i diplomatike 1. Moscow & St Petersburg: Al'ians-Arkheo.

Bischoff, B. 1979. Paläographie des römischen Altertums und des abendländischen Mittelalters. (= Grundlagen der Germanistik, 24, ed. by H. Moser & H. Steinecke.) Berlin: E. Schmidt.

Brown, M. P. & P. Lovett. 1999. The Historical Source Book for Scribes. London: The British Library.

Dobiaš-Roždestvenskaja, O. A. & W. W. Bakhtine. 1991. Les anciens manuscrits latins de la Bibliothèque publique Saltykov- Ščedrin de Leningrad VIIIe-début IXe siècle. Paris: CNRS.

Elagina, Natalia et al., eds. 2001. The Insular Gospels of the 8th century in the Collection of the National Library of Russia, Saint Petersburg: Electronic edition of the manuscript Lat. F.v.I.8. CD-ROM. St Petersburg: The National Library of Russia.

Kendrick, T. D. 1938. Anglo-saxon Art to A.D. 900. London: Methuen.

Kilpiö, Matti & Leena Kahlas-Tarkka, eds. 2001. Ex insula lux. Manuscripts and Hagiographical Materials Connected with Medieval England. Helsinki: Helsinki University Library.

Kockelkorn, R. 2000. Evangeliorum quattuor codex Petropolitanus (Lat. F.v.I.8): das hiberno-sächsische Evangeliar in der Russischen Nationalbibliothek von Sankt-Petersburg. Luxemburg: Bibliothèque nationale.

Lindsay, W. M. 1915. Notae latinae. An Account of Abbreviation in Latin Manuscripts of Early Minuscule Period (c. 700-850). Cambridge: Cambridge University Press.

Lowe, E. A. 1966. Codices Latini Antiquiores. Part XI. Oxford: Clarendon.

Privalova, E. L. 1962. "Irlandskaya srednevekovaya illuminirovannaya rukopis v sobranii Leningradskoy Gosudarstvennoy Publichnoy bibloteki [A medieval illuminated Irish manuscript in the collection of the Leningrad State Public Library]". Sredniye veka 21: 201-211.

Tassin, R. P. & C. F. Toustain. 1757. Nouveau traité de diplomatique par deux religieux bénedictins de la congrégation de S. Maur, Vol 3. Paris.

Wright, D. 1961. "Addenda". Latin Gospel books from A.D. 400 to A.D. 800 by P. McGurk, 122-123. Paris: Erasme.

Zimmermann, E. H. 1916. Vorkarolingische Miniaturen. Berlin: Deutscher Verein für Kunstwissenschaft.

Appendix