1. Date: 1102-1113

When the MS 17 scribal team set out to make their computus manuscript, they were obliged to make choices not only about its contents, but about the form in which these contents would be arranged. Would they make a basic manuscript of tables and explanatory texts, or an "encyclopaedic" core-and-penumbra manuscript, or a new style anthology, where the computus materials were subordinated to other kinds of scientific content, such as astronomy? All three models are represented in the computus manuscripts made in England in the 11th and 12th centuries. The compilers of MS 17 appear to have selected the core-and-penumbra model, but this conclusion rests on the assumption that they were really exercising a choice, and not copying tel quel from a single exemplar. That they were indeed selecting their materials with conscious design is the argument of Background Essay: MS 17 as a computus manuscript. However, doubts concerning the date of the manuscript's compilation and the location of the scriptorium where it was written have clouded this argument. Hence these questions deserve some detailed investigation.

On fol. 3v, there is a note on the four ages of world-history, which gives the annus praesens as A.D. 1110. This note is in the hand of Scribe A, as is evident, for instance, by the three types of minuscule a in adam and natiuitate in the fourth line. Corroborative evidence that the manuscript was made around 1110 comes from the annals inscribed in the margins of two Paschal tables: a paradigm Paschal table on fols. 29r-v, with annals covering 1085-1111, and the main Paschal tables. N.R. Ker recognized that the scribe who wrote the annals up to 1111 (recte 1113) in the main Paschal tables was the same scribe who wrote the glosses throughout MS 17, that is, our Scribe A. He describes this scribe's hand as "contemporary" or "nearly contemporary" with the manuscript's "main" or "text" hand, by which, apparently, he means Scribe B.1 A close look at some of the earlier annals, those for A.D. 625-646 confirms this identification: the minuscule G and the Tironian et, for instance, are signatures of Scribe A. Cyril Hart agrees that the annalist and the glossator are one, and his designation for our Scribe A is "W".2 He observes that Scribe A/W stops writing the annals in 1043 and is replaced by another Scribe (X), identical with our Scribe B, who writes the annals for 1065, 1066, 1074 and 1080. However, his claim that yet a third scribe (Y) picks up in 1092 and continues through to 1118 is mistaken. The entries for 1092 and 1095 are also by X/B. It is only when we turn the page that we find Hart's Scribe Y (i.e. Scribe A) at work, and only up to 1113.

Hart characterizes the hand of his Scribe Y as "close to that of W" (that is, our Scribe A) "but distinguishable from it."3 In fact, the two hands are one and the same. Scribe A's three types of minuscule a, as well as his characteristic g are visible throughout, and his unique way of abbreviating quod can be seen in the annal for 1104. Scribe A also wrote the annals for 1085-1111 attached to the paradigm Paschal table on fols. 29r-v. The uniform quality of ink and script in the annals for 1085-1102 on fol. 29r, as well as the manner in which these annals are laid out, suggests that they were all transcribed at once. The solitary entry on fol. 29v for 1111 was written with a different pen, and in probably contemporary with the event recorded.

That few scholars until recently have accepted the dating clause on fol. 3v at face value is closely bound up with certain misunderstandings about this paradigm Paschal table and its annals. The first to cast doubt upon the validity of the dating clause was André van de Vyver, who argued that this clause was added, together with the medical texts on fol. 1-2, by a later hand than that which wrote out the Paschal table for 1083-1102 on fol. 29r: in other words, MS 17 was written between 1083-1102, since the scribes would hardly insert an obsolete table into a computus manuscript, while the dating clause of 1110 is a later addition.4

His revised dating of the manuscript was accepted by Heinrich Henel5, and seconded by C.W. Jones, who in an undated manuscript note preserved in the Library of St John's College argued that the whole of the first gathering is composed of fly-leaves onto which the dating clause, along with the medical texts etc., was entered after the rest of MS 17 was complete, "probably by the hand of annal fol. 29b". This is Scribe A's annal for 1111, but Jones did not perceive that this same scribe wrote the annals for 1085-1102 on fol. 29r. Moreover, Jones argues that because this paradigm Paschal table begins in 1083, it, and by extension the whole manuscript, must have been written before the 19-year Paschal cycle beginning in that year expired in 1101. Otherwise, in his view, the table would be useless.6

This argument can be refuted on two grounds. On fol. 12r of MS 17, there is a ferial letter table keyed by 28-year solar cycles for the period 1092-1624; therefore by Jones' logic, MS 17 must have been written in the first solar cycle after 1092, i.e. 1095-1119, as indeed the dating clause says it is. Furthermore, in the left margin of the AEIOV key-letter table, Scribe A has indicated where "iste cyclus" ("the present cycle") begins and ends by two sigla. The initial siglum is against the letter A in the column applicable to either year 1 or year XVI. Were MS 17 completed around 1110, "the present cycle" would begin in 1102 (year 1). The next year 1 is 1120, which is probably too late, since Scribe A's final datable appearance in MS 17 is in an annal for 1113. 1117 (year 16) is, for similar reasons, less likely than 1102.

But is Jones' logic in fact valid? As a terminus a quo, such tables are useful for establishing the date of a manuscript, for one is unlikely to find a table which begins after the time of writing. On the other hand, it is very common to find tables which begin decades, or even centuries before the manuscript in which they are housed was compiled. After all, the main Paschal tables which begin on fol. 139r of MS 17 commence in A.D. 532. Henel observes that dating computus manuscripts by the initial year of their cycles is a doubtful practice, since the tables were often copied from an exemplar and then extended up to or beyond the copyist's time. Paschal tables which start at the beginning of a 19-year cycle are especially suspect; those which commence in the middle of a cycle have a slightly stronger claim to be considered evidence for dating a manuscript.7 The table on fol. 29r-v of MS 17 begins at the beginning of a cycle. Furthermore, it is important to bear in mind that this is not a functional Paschal table, but an instructional paradigm, a lesson on how such a table could be constructed and read. It is preceded by an explanatory text which deals with just these issues, and it forms part of a logically constructed anthology of texts and tables explaining the Paschal cycle. Considering this purpose, it probably did not matter much what cycle it used. A group of texts and a paradigm Paschal table identical with MS 17's appears in the Peterborough computus (Tiberius C.I fols. 2r-3r), but though the manuscript was written around 1120, the table covers 988-1006. But in fact, MS 17's compilers did make a deliberate choice to start the paradigm Paschal table in 1083 -- not because the manuscript was being written during that cycle, but because that was when Abbot Gunter arrived at Thorney Abbey.

Jones' and van de Vyver's doubts arose from failure to recognize Scribe A. Had they done so, they might still have argued that MS 17 was written, or at least begun within the cycle 1083-1101, but they probably would have placed it closer to the end of the cycle. Neil Ker has laid down a very helpful principle that "The clearest evidence <for dating> comes from annals and chronicles which look as if they have been written in one hand and at one time up to a certain date and then from year to year at short intervals."8 The annals on fol 29r could not have been composed before 1087, for the annal for 1086 refers to Gunter's appointment by the elder King William. However, the annals which cover the years 1085-1102 appear to have been transcribed all at once. They have the same even quality of ink, script, and spacing, and are altogether too tidy to have been built up year by year. Moreover, the first annal (1085), unlike the others, is not written next to the year in question, but far up the margin, and linked by a signe de renvoi, as if Scribe A foresaw that its length would leave him insufficient room for the others he had to copy out. In short, it is probable that all the annals on fol. 29r were written down after 1102. We may take this as an approximate terminus a quo for the production of J, for if the table had been available for annals at an earlier date, it would doubtless have been put to use then.

To establish a terminus ad quem, we should bear in mind that Scribe A is the guiding intelligence of MS 17. There are several parts of the manuscript that are incomplete, and that would have been his responsibility (e.g. the rubrics for Helperic's De computo, and a number of tables and diagrams, such as those on fols. 8r, 28r, 30r). That no one else was prepared to complete them argues that Scribe A's disappearance effectively brought the production of MS 17 to an end. The last annal in the great Paschal tables written by Scribe A records the death of Abbot Gunter in 1113, which accords with the dating clause on fol. 3v.

We can confirm and refine this statement in some measure by examining the work of Scribe A on the annals in the margins of the great Paschal tables. This work has two distinct aspects: annals from 532 up to his own day which Scribe A must have transcribed from elsewhere, and annals pertaining to his own time, which he may have composed as well as written down. He would appear to have worked on the two projects at the same time, for his retrospective transcription breaks off at 1043, and his colleagues filled in the gap between 1065 and 1098, when Scribe A begins again.

Cyril Hart has argued very convincingly that the retrospective annals copied by Scribe A and his colleagues derived from annals composed in Ramsey Abbey, at least up to 1081.9 In fact the annals relating to Ramsey extend to 109510; Hart overlooked this fact because he believed MS 17 was written in the 1080's; he further posited that MS 17 not only used Ramsey materials, but was actually written at Ramsey.

Scribe A, then, had Ramsey annals to copy up to 1095. His plan seems to have been to transcribe the Ramsey annals, and then insert important Thorney notices in the appropriate places. For example, against the year 972 he entered, in a slightly smaller hand but with a prominent initial capital, a brief account of King Edgar's donation of a foundation charter for Thorney to St Ethelwold. The obit of St Ethelwold s.a. 983 may be another interpolation of this type. Meanwhile, beginning on fol. 81v of Cotton Nero C.VII, Scribe A wrote down contemporary annals. These fall into two groups: short notices in the side margin, set against the year in question, and a series of more discursive annals, some covering years for which there are also marginal notices, at the top of the page, keyed by signes de renvoi. The marginal annals begin in 1102; significantly, that is the year in which Scribe A's block of annals in the paradigm Paschal table on fol. 29r end, and in fact, the identical obit for "Hugo monachus" appears in both places. The discursive annals, at least up until 1104, seem to have been composed after the events they relate. In the first annal (1098), the scribe tells of the solemn translation of relics to the half-finished church at Thorney, but admits that he cannot recall the names of all the dignitaries present; he also says that Peterborough was at that time ("erat...tunc") without an abbot. Moreover, the annal for 1104, recounting the death by drowning of a monk and several of the abbey's servants, also mentions the alms offered yearly in their memory. Even the briefer marginal entries, at least those before 1109, may not be exactly contemporary: the comet of 1106, and the death of Anselm in 1109 are both mis-dated. In connection with the year 1109, it is worth noting that in the Tabula Dionysii on fol. 30r, this year is flagged with three dots.

In short, there is sufficient evidence to support the assertion that the writing of original Thorney Abbey annals onto fol. 29r of MS 17 began no sooner than 1102, and onto the great Paschal tables probably about 1109. MS 17 must have been written between 1102 and 1113, the date of Scribe A's last annal; this validates the dating clause on fol. 3v.

2. Location: Thorney Abbey

Confirming the dating clause is important, because a date of twenty or thirty years earlier has long been an accepted article of scholarly consensus concerning MS 17. So has a certain scepticism about its place of origin. It is Cyril Hart who has most aggressively championed Ramsey as MS 17's birthplace, and many have accepted this ascription. A corner stone of Hart's reasoning is that since the annals in the great Paschal tables up to at least 1081 originated at Ramsey, therefore MS 17 must have been written at Ramsey.11 All grounds for such an assertion evaporate before a recognition of Scribe A, for if he composed, or even merely transcribed original annals recounting the affairs of "nos thornenses" (s.a. 1098), he is evidently a Thorney Abbey scribe.

Moreover, as T. M. Bishop observed, "where a manuscript is preserved, there it had been written: the presumption, sometimes nullified by better evidence, amounts to fair probability where the manuscript comes from the library of a Benedictine house already in existence at the time of writing."12 Hart's "better evidence" was that Thorney lacked the intellectual, calligraphic and artistic resources to produce such a volume as MS 17.13 Others, however, have not thought it inconceivable that a de luxe manuscript could be made at Thorney; arguments have been made that that the first abbot of Thorney, Godeman of Winchester, may have been the scribe of the Benedictional of St Aethelwold,14 or that Thorney may have been the birthplace of the "Missal" of Robert of Jumièges15 or the Gospels of Judith of Flanders.16 None of these suggestions has won general approval, but not because Thorney was considered inherently improbable. Fortunately there is more solid evidence of the talents of the Thorney scriptorium in the form of surviving manuscripts contemporary with MS 17.17

Of these, Edinburgh National Library Advocates 18.6.12 and 18.7.8 are the most instructive. These two manuscripts,18 written about A.D. 1100, originally formed one volume, and are of considerable interest in that many of the leaves are palimpsest, some in Anglo-Saxon hands of the eighth century. One of St Ethelwold's endowment gifts to Thorney Abbey was the relics of Benedict Biscop, founder of Wearmouth-Jarrow, but only further research will determine whether these might be membra disiecta of codices from that famous foundation. What is of interest to the present inquiry is the upper writing. This literary anthology contains Persius' Satires, Avianus, Abbo of St Germain, short satires, epigrams and enigmas, and Cicero's Cataline Orations. It is certainly an elegant piece of work, adorned with calligraphic majuscules and a few foliage initials, like those found in MS 17, if somewhat more florid. Unlike MS 17, the literary anthology's display capitals are frequently filled, and silver is used in the large initials. The manuscript was definitely made at Thorney, for the hand of the satirical epitaph on fol. 35v of Adv. 18.7.8 is that of Scribe A (ill. 1), and it is very likely that the main scribe of Adv. 18.6.12 was the same scribe who composed the annal for 1084 in MS 17.

Plainly, the scribal and artistic resources necessary to produce MS 17 were not wanting at Thorney Abbey in 1110, as evidenced not only by the literary anthology, but by other manuscripts surviving from this period, such as Bodleian Library Laud Misc. 364 (Scribe B can be seen at work on fols. 80v-82r) and the liber vitae of Thorney, British Library Add. 40000. We need not hesitate to ascribe MS 17 to the scriptorium of Thorney Abbey. Not only were the scribal and artistic resources available there, but more importantly, MS 17 was not created in a vacuum. It was part of a program of book production, one apparently sparked by the arrival of Abbot Gunter. The surviving products of this program witness to a fairly distinctive "house style" characterized by a bright, open, neat and slightly old-fashioned Caroline text hand, set off by crisp and simple calligraphic capitals. That house style shines out of the folios of MS 17.

1 Ker 1938, 131; Ker 1957, 435.

2 Hart 1970, 33-34.

3 Hart 1970, 34.

4 Van de Vyver 1935, 144 n.2.

5 Henel 1934, 297-298.

6 Jones' note, a copy of which was kindly furnished to me by the Librarian of St John's College, goes on: "But the construction of folio 29a shows codex constructed in dec.<ennouennal> cycle A.D. 1083-1101 and if...Coxe is right that the annals in margin fol. 29a are in a later hand, then it seems probable that this MS was written A.D. 1083-1085." Cf. Coxe 1852, 2.6, which refers to the annalist's hand as "aliquantulum recentior." Coxe was mistaken in this, but at least he recognized that the annals on fol. 29r and that which in on 29v are by the same hand.

7 Henel 1934, 21-24.

8 Ker 1960, 21.

9 Hart 1970, 35.

10 Note the annals for 1092 (accession of Alwinus as abbot of Ramsey) and 1095 (death of Wulfstan of Worcester; Worcester had very close historical associations with Ramsey). Hart did not recognize that the hand which wrote these notices (Scribe B) also wrote those for 1066, 1075 and 1081.

11 Hart 1970, 34; re-affirmed with more elaborate argumentation (notably that the Thorney scribes actually worked at Ramsey Abbey) in Hart 2003. Hart was not the first to suggest Ramsey as MS 17's place of origin: Singer 1919-1920, 372, states that it was made at Ramsey, though without giving reasons. This view was accepted by Kauffmann 1975, 57-58, and Cameron 1983, but rejected by Cameron 1993 in favour of Thorney; likewise, it was accepted by Baker 1982, but rejected by Baker and Lapidge 1995. See also Gneuss 2001, [page].

12 Bishop 1971, xv.

13 Hart 1970, 29-31.

14 Warner and Wilson 1910, xii-xiv; Wormald 1959 , 10.

15 Atkins 1928, 219 sq.

16 On the East Anglian origins of this MS, Harssen 1930, 1-13. Croyland and Thorney have been suggested as places of origin: Temple 1976, 23.

17 Surviving Thorney manuscripts are listed in Ker 1964, 189.

18 Discussed by Cunningham 1973, 84-85 and 88-89.