SOURCE: Thomson, Rodney M. “William of Malmesbury's Carolingian Sources.” Journal of Medieval History 7, no. 4 (December 1981): 321-36.

[In the following essay, Thomson examines William's historical methods through an examination of one of his sources, also commenting in general on twelfth-century historiography.]

William of Malmesbury's Gesta regum, completed in 1125, is of course primarily a history of England; primarily, but not solely, and certainly not in any narrow sense.¹ On the contrary, William felt it necessary to at least summarize, in a series of digressions, the history of those people who, by invasion, intermarriage or diplomatic intercourse, became part of England's history. So, he dealt with the continental Saxons and Scandinavians briefly, the Frankish, German and French royal families and the Normans in greater detail.² The excursuses became more frequent as, on nearing his own time, his vision grew even more pan-European, encompassing the First Crusade, the later stages of the Investiture Controversy and the appearance of new monastic Orders.³ More frequent also become the notorious and baffling fables and folk-tales, apparently (but only apparently?) introduced as light relief.⁴ William justified these excursuses in various ways. Only one of them claims our attention here, and that is his prolonged treatment of the revived Empire of Charlemagne and his successors. The subject attracted him, not merely because of his perception of the various links between these monarchs and the English ones; he was also interested in the notion of ‘empire’, which he saw as an entity with an unbroken existence stretching from classical antiquity to his own day. Its history is summarized and interpreted by him in one of his best-known collections, the partly autograph Oxford, Bodleian Library MS. Arch. Seld. B. 16, written in 1129.⁵ This double interest is made abundantly clear in the introduction to his first section on the Franks in the Gesta regum (Stubbs 1887:69):

… quoniam ad id locorum uenimus ut Karoli magni mentio ultro se inferret, uolo de linea regum Francorum, de qua multa fabulatur antiquitas, ueritatem subtexere; nec multum a proposito elongabor, quia progeniem eorum nescire dampnum duco scientiae, cum et confines nobis sint, et ad eos maxime Christianum spectet imperium.

There follow three pages on the history of the Frankish royal house, from its misty origins to Charlemagne, and thence to the end of his line and the resuscitation of the Empire by Otto the Great, from whom modernus Henricus … lineam trahit (Stubbs 1887:72). In his next book, prompted by the meagre excuse of Ethelwulf's marriage to Judith, daughter of Charles the Bald, William intrudes a further seven pages De successoribus Caroli Magni, from Louis the Pious to Louis the Child (Stubbs 1887:110-7). Both sections are based mainly upon annalistic and genealogical sources.

As Bishop Stubbs demonstrated, in his treatment of the Carolingians William is not at his most reliable, and his account contains errors and confusions (Stubbs 1889:xxi-xxvii). Stubbs managed to identify some of William's sources, but confessed himself in the final analysis beaten: “Our author's habit of paraphrasing or reproducing the matter borrowed from the earlier authorities, renders it impossible in every case to refer his statements to their primary source; nor is the amount of information which he adds to the current foreign history of early times sufficiently important to make it worth the trouble” (Stubbs 1887:69 note 2). The sources of William's information about the Franks are indeed difficult to identify for the reason Stubbs gives, but they are worth some trouble, not indeed for any great gain in our understanding of ninth-century Francia, but, because of the light shed on William's historical method.

Fortunately we are not dependent upon the Gesta regum alone for our knowledge of William's Frankish sources and his treatment of them. There exists a manuscript which throws unsuspected light upon the Frankish sections of the Gesta regum and upon William's technique as a historian. Until 1977 it was London, Sion College MS. Arc. L. 40.2/L. 21, and while there was fully described by N. R. Ker (1969), although he did not succeed in identifying all of the items in it. It is now Oxford, Bodleian Library MS. Lat. class. d.39 (hereafter referred to as B). Written almost entirely in one expert English hand of about 1175, the manuscript has no indication of provenance. However, as Dr Ker noted, a book with the same main items of contents in the same order appears in the library-catalogue of Christ Church, Canterbury, compiled by Prior Henry of Eastry at the beginning of the fourteenth century (James 1903). Bodleian Library MS. Lat. class. d.39 could be the same book, although it does not bear a Canterbury press-mark.

It contains Suetonius' Vitae Caesarum (fos. 1-125), a Genealogia regum Francorum from Faramund to Pippin II (fos. 125-125v), the Annales Mettenses priores from Pippin II to the death of Pippin III (fos. 125v-136v), Einhard's Vita Karoli (fos. 136v-150v), another genealogy from St Arnulf to Lothar I (f. 150v), the Visio Karoli Crassi (fos. 150v-152v), a chronicle of the Frankish kings from the sons of Louis the Pious to the extinction of the west Frankish line (fos. 152v-153), a florilegium of twenty-eight extracts from Aulus Gellius' Attic nights (fos. 153-9), and, in another but similar and contemporary hand, the poem Cesar tantus eras (fos. 159-159v). William knew Suetonius, Einhard, the Gellius florilegium and Cesar tantus eras,⁶ and we shall see that he also knew the Metz Annals and Visio Karoli. None of these works was very common in twelfth-century England; the florilegium and poem in particular merit further comment. It is certain that both William and John of Salisbury used the Gellius florilegium copied (incompletely) in B, and I have elsewhere advanced reasons for thinking that William might have been its compiler (Thomson 1976; Martin 1977). At all events he excerpted from it into his own much more copious florilegium, the Polyhistor (BL MS. Harl. 3969; Cambridge, St John's College MS. 97). As for Cesar tantus eras, so popular on the continent, William's text in the Gesta regum is strikingly different from all other known versions in its readings and order of lines (MGH Poetae 4.2:1074-5). It is therefore significant that the B text (not used in the Monumenta edition) follows William's version but for four variants: tollit (with three late German manuscripts) for William's tollat, ac (a scribal slip) for his ad, lenis (with Vatican Library MS. Vat. lat. 3827) for his laetus, and numina, incorrectly with seven Gesta regum manuscripts against lumina of the remainder.⁷ The seven manuscripts with the defective reading represent all three editions of the Gesta, and thus the variant probably arose from a change of mind by the author himself. Numina we may conjecture as his first choice, the n soon after cancelled with a sublinear point and the l written over it, William's usual practice as revealed by the autograph of his Gesta pontificum.⁸ A provisional explanation of these interesting facts might be that B's scribe or compiler had access to some of William's books, and perhaps copied Cesar tantus eras from the Gesta regum itself.

But the manuscript bears evidence of a more intimate connection with William than this. Scattered throughout it are marginal notes in the hand of the main scribe of the text. They were undoubtedly copied from his exemplar, and since they are certainly by a single author, they show that the exemplar contained most if not all of the items in B. In other words, B is a copy of a single manuscript with the same contents. The notes are as follows:

1. f.17 Verbum Ciceronis est adulantis Cesari in oratione pro M. Marcello: Itaque illam tuam preclarissimam et sapientissimam uocem inuitus audiui; satis te diu uixisse uel nature uel glorie. Satis si ita uis, naturae fortasse. Addo etiam si placet, glorie. At quod maximum est, patrie certe parum (Cicero, Pro Marcello, 25).

2. f.79 Verba Orosii sunt: Verum hec Suetonius dixerit hec [sic] Claudium egisse, quia CX tumultuantes Iudeos coherceri et comprimi iusserit, uel etiam Christianos uelut cognate religionis homines uoluerit expelli, nequaquam discernitur.

3. f.80 Hic est Felix qui in Actibus apostolorum legitur apostolum Paulum eripuisse a Iudeis.

4. f.139v Ex Albini epistola: Antiqui Saxones et omnes Fresonum populi, instante rege Karolo premiis et minis, ad fidem Christi conuersi sunt (Alcuin, Ep.7).

5a. f.140v Ex A(lbini) epistola: Duces regis mille partem Hispaniae tulerunt a Sarracenis, quasi ccc milia in longum per maritima (ibid.). 5b. Ludouicus etiam filius eius Hispanias ingressus, Barcilleam ciuitatem que rebellabat iterum subegit (Hugh of Fleury; Rotendorff 1638:173).

6. f.141v Rex Karolus cum exercitu irruit super Sclauos … eosque subegit (Alcuin, Ep.7).

7. f.142 Auari quos nos Hunos dicimus … inuaserunt (ibid.).

8. f.143v Greci cum classe … ceteri fugerunt (ibid.).

9. f.144 Ex decretis: … (Ivo of Chartres, Panormia 8. 135).

10a. f.147 Ex Haimonio: Sub eius imperio sunt multa et precipue he monasteria facta … piissimus imperator insigniuit (Hugh of Fleury; Rottendorff 1638: 175). 10b. In Gallia monasterium sancti Pauli Cormaricense de medietate possessionum quas habebant canonici de sancto Martino, hortante Alcuino, unde hodieque pro anima ipsius Alcuini fiunt ibidem elemosina multa cotidie.

11. fos.147-147v Ex Haimonio: Cognati Adriani pape … anno incarnationis dcclxxxiiii (ibid. 177).

William wrote this sort of note, amplifying the main text from other sources, in his various collections, for example in the already mentioned Bodleian Library MS. Arch. Seld. B. 16 (Thomson 1975:379, 383, 385; 1978:390, 409-10). I believe that he wrote the originals of these notes too. Let us give our attention to the sources that they quote. Orosius and the book of Acts William and many another knew.⁹ But Cicero's Pro Marcello is a different story. The three Orationes Caesarianae (Pro Marcello, Pro Ligario, Pro rege Deiotaro) travelled together, and they are included in the late copy of William's great edition of Cicero's speeches and philosophical writings.¹⁰ There is some doubt as to whether they were not added to the collection after William's time, and I have not found any citation from them among his other works (Mynors 1963). Still, William was the greatest English Ciceronian of the twelfth century; the Orationes Caesarianae had some circulation, and one would expect him to know them.¹¹ More convincing are the four notes from Alcuin's letter 7. This letter is only transmitted in two early eleventh-century English manuscripts, and the readings of these extracts point to one of them, British Library Cotton Tiberius Axv, probably from Christ Church, Canterbury.¹² William knew this manuscript, made a copy of it, and extracted from it in the Gesta regum and Gesta pontificum (Thomson 1977:147-61). In fact, he quoted letter 7, including the passages given in notes 4, 5a and 6, in the Gesta regum (Stubbs 1887:92). Hugh of Fleury, quoted in notes 5b, 10a and 11, William used in the same work,¹³ and he summarized his Historia in Bodleian Library MS. Arch. Seld. B. 16.¹⁴ There he calls him Haimo, as in notes 10a and 11 (although they seem to suggest the form Haimonius). In fact all three extracts are substantially as in the Selden manuscript, f.139, where the last is accompanied by a reference to Einhard, against whose text it is written in B.¹⁵ We come to note 9, presenting an extract from a letter fraudulently ascribed to Pope Hadrian, exactly as given in Ivo's Panormia. The Panormia was introduced into England about 1100, and rapidly became popular.¹⁶ In the second half of the twelfth century there was a copy at Cirencester, not far from Malmesbury (Oxford, Jesus College MS. 26), and a house with which William had had some contact.¹⁷ He knew a number of the bishop's writings, and in the Gesta regum gives a resumé of this very extract, which he puts into the mouth of Pope Gregory VI (Stubbs 1887:250-1). Finally, there is the curious note 10b, concerning Alcuin's patronage of Cormery, a dependent cell of St Martin's abbey actually founded by his predecessor Abbot Itherius.¹⁸ The erroneous description of the St Martin's community as canonici, and the mention of the daily alms, suggest that the basis of this note was a late and presumably oral communication. A variant of it is found in some manuscripts of the Gesta regum (Stubbs 1887:lv and 69 note 1), either in the text or as a marginal note—the latter doubtless reflecting the situation in William's autograph. His note runs Iacet in Francia apud Sanctum Paulum de Cormarico, quod coenobium Karolus magnus eius consilio construxit; unde hodieque quatuor monachorum uictus et potus pro eiusdem Alcuini anima quotidianae infertur elemosinae in eadem ecclesia. This note, too, contains erroneous information, namely that Alcuin founded Cormery and that he was buried there. None of the contents of either note is found in any other known written source. On the other hand, each note has specific information not in the other, excluding the possibility of either being adapted from the other. They are best accounted for as the work of one man, each reflecting with different emphases information which he carried only in his head. It is also perhaps worth noting the use in both of hodieque for hodie, a peculiarity of William's.

My earlier suggestion, that B derives from a group of William's books, must consequently be modified. Rather, it was copied from a single one compiled or acquired for or by William and bearing his annotations. With this in mind, it is time to turn to the Carolingian chronicles in it, and their relationship to the similar material in the Gesta regum. These chronicles make up a compilation which, as the numerous cross-references show, was the work of one man. It begins with the conflation of an otherwise unknown regnal list and genealogy:

(f.125) Incipit genealogia regum Francorum. Anno ab incarnatione Domini CCCXXIIIJ regnauit primus rex Francorum Faramundus annis v. Successit filius eius Clodio annis vi. Filius Clodionis Meroueus annis xiiii. Ab hoc omnes reges Francorum Merouingi uocati sunt. Childericus filius Merouei regnauit annis xxiiii. Clodoueus filius Childerici annis xxx. Hic primus fidem Christi suscepit. Lotarius filius Clodouei regnauit annis li. Chilpericus filius Lotharii annis xxiiii. Lotarius filius Chilperici annis liiii. Dagobertus filius Lotharii annis xxxiiii. Clodoueus filius Dagoberti annis xvi. Clotarius filius Clodouei annis iiii. Theodoricus filius Lotharii annis xviii.

Medio tempore Botgisus illustris (f.125v) dux Francorum ex filia Chilperici regis filii Lotharii primi genuit Arnulfum sanctissimum Metensis urbis episcopum. Arnulfus genuit Angisum. Angisus genuit Pipinum, de quo ita historia narrat.

(Here begin the Annales Mettenses priores with a new line and painted initial.)

The regnal list distantly resembles one published by Holder-Egger from Le Havre MS. 1 of the eleventh century (Holder-Egger 1890). There is also a slight resemblance to some early entries in the Rouen Annals (Prévost 1855; Liebermann 1879). These parallels, although not close, emphasize that the list is probably dependent upon a late source, and certainly upon a corrupt and confused one. Botgisus or Botgisel is given as St Arnulf's father in a genealogy discussed by Bonnell and copied later on in B, but this example is unique in making his wife Chilperic's daughter rather than the daughter of his son Lothar. It too is probably based upon a late and unauthoritative source. Neither the list nor this genealogy were used in the Gesta regum. There, William reproduces a quite different genealogy for the Arnulfings, the so-called Domus Carolingiae genealogia, in a redaction made at the abbey of St Wandrille (Stubbs 1887:71 lines 1-8; MGH Scriptores 2:308-9). However, his statement about Meroveus, a quo omnes post eum reges Merouingi sunt, has an analogue in the regnal list, and his date for Faramund's accession is only a year later than its 424 (Stubbs 1887:70). More important than the slight discrepancy, which could easily be due to a scribal slip (of v for iv), is the fact that, so far as I am aware, no other source assigns Faramund a date remotely similar to these.

Then follow the Annales Mettenses. Until the year 717, they are rendered verbatim except for small omissions which might have been accidental. In the course of the entry for 717 summarizing begins, with some unfortunate results. The events of 725 and 731 were somehow telescoped under 721, and the years 732-746 misread as 722-736. The entry for 738 was omitted altogether. In the margin of f.133, opposite the entry for 731 (741), is a note by the scribe, beginning with a coloured initial as the entries in the text: Anno Domini DCCXXVIII Aquitanos Karolus iterum tributarios fecit. This is based on the Annales Mettenses entry for 738, except that Aquitanos should read Saxones. Heavy summarizing begins at 754,¹⁹ and continues to 768, where the extract ends. This last, summarized section reads as follows:

(f.136v) … et [Pipinus] Papiam obsedit, usquequo ipsum Haistulfum ad deditionem compulit, et se omnia redditurum Romanis que eripuerat sa (f.137)-cramento firmare fecit. Sed reddeunte Pipino Franciam, Haistulfus spreto sacramento nichil Romanis reddidit, sed insuper Roman graui obsessione afflixit. Quo per litteras pontificis cognito, Pipinus iterum Italiam aduolat, iterum Haistulfum ad redditionem ereptarum ciuitatum in scripto compellit. Quam redditionis cartam excellentissimus Pipinus per Fulradum uenerabilem abbatem et consiliarium ipsius Romam Sancto Petro cum clauibus ciuitatum misit. Post non multum uero tempus Haistulfus in uenationem pergens diuino nutu percussus est, cui Desiderius quidam successit, quem postea Carolus magnus a regno expulit, sicut sequens liber declarabit. Pipinus uero reuertens Franciam xiii. annis postea uixit, et regno Francorum undique dilatato ix. kal. Octobris obiit, et apud Sanctum Dionisium sepultus est.

It is a surprise to find no textual relationship whatever between this and the only other known English version of the Annales, the early twelfth-century copy from Durham (Durham, Cathedral Library MS. C. iv 15; Simson 1905:v-vi). Rather, this version shows a close connection with the important late twelfth-century manuscript from St Arnulf's Metz, now Berlin, MS. Lat. 141 (Rose 1893; Simson 1905:ix-xi; Grat 1964). This manuscript presents the Annales Mettenses iuniores, a conflation of the Annales priores in a form already somewhat abbreviated, and the Chronicle of Regino of Prüm. The B version drew on the same slightly abbreviated form of the Annales, but is not interpolated from Regino.²⁰ It thus depends upon the same manuscript of the Annales as the Berlin text, and that manuscript could hardly be other than a Metz one. Let us now return to William. Could he have known the B version of the Annales Mettenses? In the first Frankish section of the Gesta regum, the account of the rise of the Carolingians seems certainly based upon the Annales.²¹ Skeletal though it is, the sum total of its information is found in the Annales and only there (in particular the date of 687 for Pippin II becoming count of the palace), and its emphases reflect those of the relevant section of the Annales. It is, then, surely significant that William's use of the Annales in the Gesta regum is confined to the years covered by the B extract, that is, the period up to the death of Pippin III. Thereafter, his main source is Einhard, whose text follows the Annales in B.²² It is also worth noting a curious mistake, unusual for William, in his account of Pippin III's reign, where he refers to Desiderius, meaning Aistulf (Stubbs 1887:71). Supposing William to have used the B version of the Annales makes it possible to see how he might have made this slip. It would certainly have been easier to make from a cursory glance at the summary rather than the full text for the relevant events.

Next in B comes Einhard, much used by William in the first Frankish section of the Gesta regum, and then the following genealogy:

(f.150v) Prosapia regum qui a beato Arnulfo usque in hec tempora geniti constant, uel quot tempora regnum Francorum ampliauerunt, uel gubernauerunt, his annotata habentur. Botgisus illuster uir genuit Arnulfum, sanctissimum uirum episcopum urbis Metensium. Beatus quoque Arnulfus genuit Angisum. Angisus genuit Pipinum maiorem domus. Pipinus genuit Karolum ducem. Karolus genuit Pipinum regem, Pipinus quoque rex genuit Karolum regem famosissimum et primum regem Francorum et imperatorem Romanorum. Karolus imperator genuit Ludouuicum imperatorem qui regnauit annis xxvi. Hludouuicus imperator genuit Ludouuicum et Karolum et Hlotharium imperatorem, qui nomen quidem imperatoris annis xv. obtinuit sed monarchiam regni cum fratribus sortitus est. Lotharius genuit Ludouuicum qui regnauit […].

With the exception of certain interpolations, distinguished by italics, this reproduces a well-known genealogy discussed by Bonnell and others.²³ For the Monumenta edition, two manuscripts were used, in both of which this genealogy followed a text of Einhard, as here. Thus, a manuscript of Einhard plus this genealogy was one of the physical constituents of the Frankish compilation in B. At this point, too, one should note the Metz interest shared by this genealogy, the earlier one and the Annals. Probably all of these and Einhard were extracted from a single manuscript of Metz provenance.

Next comes the Visio Karoli, an account of the emperor's alleged vision written at or near Rheims about 900, which William copied verbatim in the second Frankish section of the Gesta regum.²⁴ Unfortunately, this work survives in a great many manuscripts, and as most have not been collated, no sure deductions can be made about the relationship between this copy and the one in the Gesta regum. Nevertheless, it must be significant that there are only four variants between the two, all of which seem to represent stylistic ‘improvements’ made to the Gesta version, according to William's usual practice.²⁵ The small number is particularly noteworthy because it has been demonstrated that B's scribe tended to make slips (Thomson 1976:384-5). Yet even so, there is no reason why the Gesta text could not have been derived from B's exemplar.

We come to the last and most perplexing section of the Carolingian compilation in B. Like the Frankish sections in the Gesta regum, it is a pastiche of which the elements are not easily isolated. It will be simplest to give it in extenso with commentary.

(f.152v) Post K (arolum) magnum ut supradictum imperauit in omni imperio patris Ludouicus annis xxvi.

The date is as in the genealogy printed above, where it is apparently an interpolation.

Post eum diuiserunt filii eius regnum. Et Karolus quidem in ea parte Gallie que nunc proprie Francia dicitur, et in Britannia minori et in Flandria, et in Normannia, et in Andegauiensi prouintia et Aquitania et Wasconia et Burgundia regnauit, et diuisit eas ducibus suis sicut usque hodie diuise sunt.

The italicized words recall part of the socalled Continuatio Adonis, another St Wandrille product, copied verbatim in the second section on the Franks in the Gesta regum:²⁶

Carolus uero medietatem Franciae ab occidente, et totam Neustriam, Brittanniam, et maximam partem Burgundiae, Gothiam, Wasconiam, Aquitaniam. …

In the B passage, however, those geographical names unintelligible to the twelfth century (Neustria, Gothia, medietatem Franciae ab occidente) have been updated or omitted, and the nationality of the compiler and his audience indicated by the addition of minori to Britannia. The next passage introduces a new source:

Ludouicus alter L (udouici) filius regnauit in Saxonia et in omni Germania preter aliquam partem que iacet iuxta Alpes. Hanc enim tertius filius Ludouici accepit Lotharius, et ex nomine suo Lotharingiam quasi Lotharii regnum uocauit, et totam Italiam cum imperio Romanorum. Regnauit xv. annis, et monachus factus in extremo uite migrauit.

The passage et ex-uocauit recalls the Historia Francorum Senonensis (MGH Scriptores 9: 367), Regnauitque-migrauit Ado of Vienne, used by William in the Gesta regum.²⁷ But the whole section resembles the first nine lines of the second Carolingian excursus in that work (Stubbs 1887:110):

… quorum Lotharius … regnauit annis quindecim in ea quae iacet iuxta Alpes parte Germaniae, et nunc Lotharingia, quasi regnum Lotharii, dicitur, et in tota Italia cum Roma. In extremo uitae … seculo renuntiauit …

The compilation in B continues:

Post eum filius eius Ludouicus eodem annorum numero [eius filius Ludouicus iiii. annis interlined by the scribe], ea potestate functus est. At uero Karolus frater Lotharii ut superius dixi rex Francie optime litteratus fuit. Quo poscente, Iohannes Scottus Ierarchiam Dionisii de Greco transtulit in Latinum. Sed et Anastasius bibliothecarius Romanae sedis ad personam eius fecit glosas eiusdem Ierarchie, sed et Passionem beati Demetrii.

The source of information for the passage Quo-Ierarchie would be a standard collection of John's Dionysian translations, with Anastasius' glosses and prefatory letter addressed to Charles the Bald. William possessed just such a manuscript (MGH Epistolae 7:430-4, no. 13; Thomson 1979). In the Gesta regum and in the letter to Peter prefixed to his manuscript of John the Scot's Periphyseon, William recalls some of this passage: [Regis] ergo rogatu Ierarchiam Dionisii de Graeco in Latinum … transtulit (Stubbs 1887:cxliv, 131). The knowledge that Anastasius also translated a Passio Sancti Demetrii for Charles the Bald could only have come from the author's prefatory letter (MGH Epistolae 7.5:438-9). This rare item survives in only two manuscripts, Brussels, Bibliothèque Royale MS. 8690-8702, and Alençon, Bibliothèque Municipale 10, both of the twelfth century, the latter from Saint-Evroult.²⁸ From such a centre it could easily have been transmitted to England.

Cuius filiam Aþulfus rex Anglie a Roma rediens uxorem duxit nomine Iuditham. Porro Ludouicus rex Germanie genuit Ka (f. 153)-rolum istum cuius uisio superius scripta est, qui cum regno patris sui accepit etiam imperium patrui, anno Dominice incarnationis DCCCLXXXIIII, et posteriori anno obiit.

The mention of Ethelwulf's marriage to Judith is found in the Gesta regum; its source is apparently Asser.²⁹ The information on Charles the Fat's assumption of the imperial title and death appears to be based on the Anglo-Saxon chronicle, although the dates have become confused (Whitelock 1961:51-2, years 885, 887). The Chronicle was used by William in the Frankish and other sections of the Gesta regum.³⁰ After this, a new paragraph opens the last section of the Frankish compilation in B, based largely, perhaps entirely upon a single source.

Ludouicus nepos Ludouici senioris apud Ferrariis qui locus est in pago Senonensi et uxor eius, nullo presente metropolitano, a quibusdam episcopis sunt consecrati et coronati. Supererant duo filii Rotberti Andegauorum comitis, qui fuit Saxonici generis uir. Senior ex his Odo uocabatur, Rotbertus alter. Ex his maiorem natu Odonem Franci et Burgundiones Aquitanicique proceres, congregati in unum, licet reluctantem tutorem Karoli pueri regnique eligere gubernatorem, quem unxit Galterius archiepiscopus Senonum. Karolus simplex in carcere positus, Radulfum nobilem filium Ricardi Burgundionum ducis, quem ex sacro fonte susceperat, una cum consilio Hugonis magni filii supradicti Rotberti et procerum Francorum, in regnum sullimauit. Vnctus uero est in regem ipse Radulfus iii. idus Iulii, Suessionis ciuitate. Post mortem Radulfi regis, per Willelmum archiepiscopum ab Anglia reductus est Ludouicus filius Caroli simplicis in Franciam et xiiii. kalendas Iulii unctus est in regem apud Laudunum. Post translationem uero regni Francorum in progenie Hugonis magni ducis, Rotbertus piissimus rex filius Hugonis regis Aurelianis consecratus est in regem. Cuius filius Hugo [meaning Henricus] iuuenis Compendii est unctus et in regem sullimatus.

The italicized passages in this section reproduce verbatim the Historia Francorum Senonensis, a chronicle of Frankish history written at Sens in 1015 or soon after (MGH Scriptores 9:364-9; Lot 1891 and 1936). The widely-spaced passages come from the same source, though not verbatim. But it is particularly interesting that several more passages not found in the printed text of the Historia also show an obvious connection with Sens, and indeed the homogeneity of the language in this section would suggest a single author or source for the whole. It is possible that we have here a fragment of a lost Sens chronicle from the early eleventh century which was used as a basis for the extant Historia Francorum Senonensis.³¹ The Historia itself was well-known in France and Normandy by about 1100, and in fact the earliest surviving manuscript of it, of about that date, has recently been identified as English.³² Even so the Sens material in B, as with the Annales Mettenses, seems an eccentric and exotic component in the repertoire of an Anglo-Norman historian.

The portion of the Frankish compilation in B which follows the Visio is thus shown to be itself a mosaic put together from a number of sources, not all perhaps identified. The manufacturer of this portion was certainly English and working in the first half of the twelfth century. This is shown by his revision of certain geographical names, already noted, by his use of Asser and the Anglo-Saxon chronicle, by his knowledge of Anastasius' Passio Sancti Demetrii prologue, known only from twelfth-century manuscripts, and of John the Scot's Dionysian translations which seem not to have reached England before about 1100.³³ Moreover, the compiler of this portion must have been the compiler of the whole, as his references to other parts of the Frankish compilation show. It is not difficult, in the light of what we have so far learned about B and its Frankish compilation, to show that the compiler was William of Malmesbury. First of all, most of the B compilation or its elements were used in the Gesta regum: some details from the regnal list, a summary of the same part of the Metz Annals, Einhard, the Visio, and, most crucially, a few passages from the concluding section, itself the concoction of an Englishman working in the early twelfth century. Conversely, texts known to William and quoted in the Gesta appear in various forms in the B compilation: Ado of Vienne, the Continuatio Adonis, a manuscript of John the Scot's Dionysian translations, Asser and the Chronicle. So far all is plain sailing, but we must take account of the fact that items used in one work are not found in the other. The maker of the B compilation omitted most of the St Wandrille material quoted in the Gesta regum, and yet, as Levison showed, this material (the Domus Carolingiae genealogia and Continuatio Adonis) was almost certainly available to him in the same manuscript which supplied him with the Visio, and towards the end of his work he included some reminiscences of one part of it, the Continuatio (Levison 1902:500 and notes). This means that he excluded most of it deliberately, in preference to the similar Metz genealogies. Why he preferred the one to the other is not as clear as why he would have avoided including both. This remains a problem whether William was the compiler or not. On the other hand, in the Gesta William shows no knowledge of the two Metz genealogies. One can understand why he would have omitted the first, which is only a summary of the second. If he did know the second, then he apparently preferred the St Wandrille alternative, again for reasons not clear to us. Perhaps it makes good sense to see William, identifiable as B's compiler, faced with a choice between similar material from two different manuscripts, and deciding to include the Metz items in one book, the St Wandrille ones in another.

My general conclusion is, then, that Bodleian Library MS. Lat. class. d.39 was copied from a book of the same contents written by or for William of Malmesbury, who provided it with marginal annotations. This I consider to be beyond reasonable doubt. The further conclusion, that he was also responsible for the compilation on Frankish history in it, is based essentially on a high degree of coincidence between texts known to William and used in the compilation, and to that extent falls short of absolute proof. Nevertheless, I think it highly likely that he was the author of the compilation, and that he put it together before 1125, since so much of it appears in the Gesta regum. We should not, however, rule out the possibility that William had only the sources of the compilation at hand by this date. On the whole, the balance of probability seems to me to marginally favour the former alternative. In comparison with his collections of Roman history, of canon law and Pope Leo's works, this one seems less mature and rather hastily constructed.³⁴ It is more akin to William's Liber pontificalis edition, also put together from heterogeneous sources, shortly after 1119 (Thomson 1978b). Like it, I would want to see this work as a product of his comparative youth. Possibly it was his engagement in this sort of enterprise that prompted Queen Matilda to request the production of the Gesta regum (Thomson 1978a:391, 410-11; Könsgen 1975).

We are now ready to tackle the important question of what William's treatment of the Franks in the B compilation and in the Gesta regum tells us about his historical technique. First of all, a historian's work is conditioned by the available sources, and one can now give a precise account of William's Frankish materials. For his general outline, he relied upon the chronicles of Ado of Vienne and Hugh of Fleury, and the Metz Annals, with some snippets from the Anglo-Saxon chronicle, Asser and perhaps some unidentified written sources.³⁵ On Charles himself he had Einhard, and for the later Carolingians chronicles from Sens and St Wandrille, the Visio Karoli Crassi and the writings of John the Scot and Anastasius Bibliothecarius. Finally, he had a regnal list and an assortment of genealogies. Moreover, we now know something of the form in which some of these works were available to him. A manuscript from Metz or a transcript of it supplied him with the Annals, two genealogies and probably Einhard, another, which Levison thought originated ultimately from St Bertin (Levison 1902:500 and notes), contained the Visio and St Wandrille items, and we should probably add another for the Sens chronicle. How long some of these books had been in England cannot be said, but I am inclined to see them as recent entrants, perhaps sought by William himself, with Norman religious houses as the main agencies of transmission. This would be consistent with other evidence for William's involvement in such a process, and we should probably imagine the books making the journey in both directions, that is coming, being copied and then returned.³⁶

This reminds us that sources are not, strictly speaking, ‘available’ to a historian: they must be recognized and a decision taken to use them. On the one hand, some of these sources were rare in England at this date, and William must have put some effort into obtaining them, whether or not he sought them from overseas. On the other hand, he was selective in the employment of what he had, both in the B compilation and in the Gesta regum. He used Alcuin's letters, for instance, in his two major histories, but to illuminate the history of post-Bedan England, not the reign of Charlemagne (Thomson 1977). The chronicles of Ado of Vienne and Hugh of Fleury, both used in the first Frankish section in the Gesta regum, hardly appear at all in the compilation. One can, I think, reconstruct the reasoning which governed this process of selection. Neither in the Gesta nor the compilation was William committed to providing the fullest possible account of the Franks. In the Gesta regum the Franks represented a deviation from his main theme, and so their history could be appropriately and briefly covered with the aid of general chronicles, not necessarily either contemporary or authoritative. The B compilation was also no more than a compendium, therefore of a skeletal, annalistic character, with the exception of Einhard's Life and the Visio. The former was surely there because of the importance of its subject; the latter William probably included because his materials for that period were so thin, and because of his lively interest in political prophecy.³⁷ But the compilation was intended to be more authoritative than the Gesta regum's asides. Thus William tried to use contemporary sources, possibly available to him in what he thought of as early manuscripts.

This labour of selection suggests that William evaluated his sources according to rational criteria. There are two particular cases which bring his critical judgement into sharper relief. Firstly, one notes that, both in the B compilation and in the Gesta regum, he discarded the Metz Annals for the period covered by Einhard's Life. The obvious reason was that Einhard provided more intimate (and therefore, for William, more weighty) information about Charlemagne's character and rule. In common with other historians of his time, moreover, he regarded literary form as one yardstick of auctoritas; a biography was, therefore, inherently more trustworthy than a set of annals (Thomson 1978a:394-5). The second case occurs in the first Frankish section of the Gesta regum, in which he makes extensive use of Hugh of Fleury's Historia for his opening description of the origins of the Frankish people. One prominent feature of Hugh's account, however, he significantly omits. This is his reference, shared with many other chroniclers from Fredegar on, to the Franks' Trojan ancestry (Rottendorff 1638:104; Wallace-Hadrill 1962). That William ignored it deliberately is beyond doubt, for this must be what he meant in his introductory passage to this section, in which he introduced the Franks, de qua multa fabulatur antiquitas, but of whom he will ueritatem subtexere (Stubbs 1887:69; Ray 1980:21). One is forcibly reminded of his famous description of Arthur, de quo Britonum nugae hodieque delirant; dignus plane quem non fallaces somniarent fabulae, sed ueraces praedicarent historiae … (Stubbs 1887:11; Ray 1980:21). Why William rejected the Trojan origin of the Franks is suggested by a slightly later passage (Stubbs 1887:70):

Nepos Faramundi fuit Meroueus, a quo omnes post eum reges Merouingi uocati sunt. Eodem modo et filii regum Anglorum a patribus patronymica sumpserunt: ut filius Edgari, Edgaring; filius Edmundi, Edmunding, uocentur; et ceteri in hunc modum; communiter uero Athelingi dicuntur. Naturalis ergo lingua Francorum communicat cum Anglis, quod de Germania gentes ambae germinauerint.

Both the Franks and English were Germanic, of the same ethnic and linguistic stock; therefore, if one hailed from Troy, so did the other, and William was obviously not prepared to accept this. He knew, from Dares and Virgil, that the descendants of the Trojans were the Romani, a race entirely distinct from the Germans.³⁸

William may have had the Gesta regum's needs in mind when he made the B compilation, but they cannot have been his whole, or even his principal justification for its manufacture. The relationship between the Gesta regum passages and the compilation is too loose for the latter to have been conceived solely as praeparatio for the former. At least two other reasons can be advanced for the making of the compilation. Firstly, as with William's other collections, it was doubtless produced as much for the edification of others, meaning in particular the Malmesbury community, as for William's own needs. But even his own needs were wider than the terms of reference for the Gesta. His historical interests were not confined to England's past, let alone its ecclesiastical heritage. He was by temperament a universal historian, who not only read all the historical works which he could find, but who liked to make up his own handbooks from them, upon topics which especially fascinated him. This he did for the history of the papacy to his own time in his papal chronicle based upon the Liber pontificalis (Thomson 1978b). He did it on the theme of ‘empire’, as mentioned earlier, in the Selden manuscript. And he did it, I suggest, for the Carolingians in the collection which we have studied here. It is worth emphasizing the breadth of William's historical vision. It has become fashionable in recent years to account for the extraordinary phenomenon of Anglo-Norman historiography in terms of a reaction by English monks to the assault upon their traditions by the Norman prelates.³⁹ William certainly registered this reaction, in his Gesta pontificum, his Antiquities of Glastonbury, and his saints' Lives. But this did not constitute the mainspring or point de départ of his historical interest and output. It conditioned only the form of some of his works, and the commissioning and patronage which lay behind them. What the mainspring was is not so easy to say. I suggest, however, that it is not to be discovered by scrutiny of William's English environment or of England's recent past. Rather, we must analyse more intently the development of historical awareness among western European literati generally during the early twelfth century.⁴⁰

Notes

1. On William himself, see the introduction to Stubbs 1887; Farmer 1962; Thomson 1978a.

2. Stubbs 1887:8-9 (Saxons); 75-6, 219 and 1889: 317-20 (Danes, but William does not seem to have had much information on their early history outside England); 1887:69-72, 110-7 (Franks); 1887:230-6, and 1889:342-4 (Germans); 1887:139-40 (French): 1887:137-9, and 1889:285-97, 320-2 (Normans).

3. Stubbs 1889:390-463 (Crusade); 322-6, 489-92, 498-509 (investitures); 380-5 (Cistercians).

4. Stubbs 1887:193-206, 231-5, 253-60; 1889: 295-7, 344-5; Ward 1976; Thomson 1978a:396-9.

5. For its contents, see Thomson 1975:387-8, but it is there dated wrongly. The evidence is set out on p. 388 note 1, and requires the correction of xlii for xl, which then produces the date given here. There is further internal evidence for the same date. For instance, on f. 140 are regnal lists in William's hand; that of the German emperors ends with Henry V, qui xx. annis post patrem regnans habuit filiam Henrici regis Anglorum, quae post mortem mariti nupsit filio comitis Andegauensis. The marriage took place on 17 June 1128. The list of kings of France ends with Louis VI (1108-37).

6. For Suetonius, Thomson 1975:379; for Einhard. Thomson 1975:388 and below, notes 15 and 22; for Gellius Thomson 1975:381-2, and 1967a; Cesar tantus eras is in Stubbs 1887:235.

7. Stubbs 1887:235 note 2; … defecta … lumina appears to be an echo of Statius, Thebais 12. 325.

8. Oxford, Magdalen College MS. lat. 172. William's corrections and alterations are faithfully recorded by Hamilton 1870.

9. William copied Orosius in Bodleian Library MS. Arch. Seld. B. 16, fos. 11-71.

10. Cambridge, University Library MS. Dd. 13. 2; Thomson 1975:372-7; Hunt 1975.

11. The stocks of John of Salisbury, his main rival in this sphere, continue to fall: Rouse 1978.

12. Citing the edition of Dümmler, E. 1894. MGH Epistolae 4. Note 4 Fresonum with BL MS. Cotton Tiberius A xv (Dümmler's A1) and Stubbs 1887: Frisonum A2; note 6 irruit with A1 and Stubbs 1887: inruit A2.

13. Stubbs 1887: most of 69 line 13 until 70 line 5 (Rottendorff 1638: 104ff); 71 line 8, the epithet of Tudites for Charles Martel (Rottendorff: 162); 72 line l, the description of Pope Leo's assailants as Hadrian's consanguinei (Rottendorff: 172); most of 110 lines 1-16 (Rottendorff: 179-80).

14. Thomson 1975:387. William used Hugh's second edition of 1110.

15. The equivalent to this note and the reference to Einhard in the Selden MS. runs: Denique ista et multa preterea monasteria sub eo facta: monasterium Sancti Philiberti, Sancti Maxentii, Concas, Murate, Magniloci, Nursiacum, Sancti Saluii, Nobiliacum, Sancti Theofridi, Sancti Pascentii, Dorosa, Solemniacum, monasterium puellare Sanctae Mariae, item Sanctae Radegundis, monasterium Deuera, monasterium Deuta in pago Tholosano, monasterium Valida in Septimania, Sancti Aniani, Galimme, Sancti Laurentii, monasterium in Rubine, monasterium Caunas. Et alia quibus ueluti gemmis tota decoratur Aquitania. Haec ideo de Carolo dicimus, omissis omnibus eius gestis quae in ipsius uita leguntur, ut ostenderemus quomodo et quantum imperatores Constantinopolitani perdiderint Romanae urbis imperium …

16. Brooke 1931. The provenances of the manuscripts listed by Brooke need to be checked against Ker 1964.

17. William and Robert of Cricklade, canon of Circencester and then prior of St Frideswide, seem to have known each other; Robert certainly knew some of William's works; Thomson 1975:393 and note 7.

18. On Cormery see Bourassé 1861; Kleinclausz 1942. Alcuin already claimed to be its founder (MGH Epistolae 4:309); compare Ardo, Vita Benedicti Aniani. MGH Scriptores 15:210.

19. The entry for 754 is reasonably intact to Simson 1905:47 line 10.

20. Specimen readings: (Simson 1905) 1 line 6 gestorum] omitted B as the Berlin MS. (Simson's B1); 3 lines 15-16 uirtutum—dubitatione] omitted as B1; 4 line Itaberga] Luberga B, compare Iuberga B1; line 8 potestatis] etatis B and B1; line 19 Osterliudos]—liudo B and B1; 5 lines 21-2 uisi sunt] fuerunt B and B1; 6 line 1 seuitia] ob seuitiam B and B1.

21. Stubbs 1887:70 lines 21-30 (Simson 1905:1-21), 71 most of lines 9-27 (Simson:21-55).

22. Stubbs 1887:70 lines 10-11, 71 lines 9-10 (Einhard, ch. 2?), 15-16 (Einhard, ch. 2), most of 71 line 27 until 72 line 13 (Einhard, chs. 3, 15, 23, 28, 30).

23. Printed in MGH Scriptores 13:246 no. 3, using Paris, BN MS. lat. 4955 and London, BL MS. Egerton 810 (about 1200, from Germany); Bonnell 1866; Hlawitschka 1965.

24. Stubbs 1887:112-16. Stubbs (1889:xxxi-ii) wrongly thought that William took this passage from the chronicle by Hariulf of Saint-Riquier; compare Levison 1902, and Poupardin 1901. William was not the only Englishman who knew it. It is summarized in the Chronicon fani Sancti Neoti, written at Bury abbey in the early twelfth century. See Stevenson 1959:102 and note 2, 140.

25. Thomson 1977:153-4. William could never resist these, even he claims, as here, to be rendering the original verbatim (Stubbs 1887:116).

26. Stubbs 1887:110-12; MGH Scriptores 2:324-5. Its title, bestowed by d'Achery and Pertz, is misleading. As far as I am aware, it is not found in any manuscript appended to Ado's Chronicle. It is an independent work.

27. MPL 123:135-6; compare Stubbs 1887:69 line 13 to 70 line 5 (MPL 123:95), along with Hugh of Fleury; 110 line 5 annis quindecim (MPL 123:135).

28. The Alencon MS. contains a recension of the canons of the Council of Clermont, 1095, related to that used by William and Orderic Vitalis; Somerville 1972.

29. Stubbs 1887:109; Stevenson 1959:9. The same information could have been obtained, with great difficulty, from the Anglo-Saxon chronicle (Whitelock 1961:43-4, 51). William was certainly using other parts of Asser at this point in the Gesta.

30. Stubbs 1887:72 lines 5-7 (Whitelock 1961:37, year 799); Thomson 1975:389-90.

31. Fliche 1909; Bautier 1971, raising the possibility of a lost work used for the Annales de Sainte-Colombe, the Historia Francorum Senonensis and Odorannus' Chronicle.

32. Ehlers 1978. Professor B. Bischoff, responsible for suggesting the provenance, dates the MS. to about 1100 or the first half of the twelfth century. From Ehlers's facsimiles, I should be inclined to make “about 1100” the upper limit for this flamboyant example of late English caroline.

33. All English MSS. of this collection seem to depend upon Oxford, St John's College MS. 128, about 1100, of unknown provenance. William's extracts from Anastasius's letter (Stubbs 1887: cxlvi) are related to this MS.

34. Bodleian Library, MS. Arch. Seld. B. 16 and Oxford, Oriel College MS.42; Thomson 1975:383-8; the same 1976b and 1978a:407-9.

35. I cannot identify the sources for Stubbs 1887:71 lines 31-3 (et ab imperatoria … temperans), 110 lines 13-14 (Carolum … exosculabatur) and 15-16 (et Wasconia), 116 line 23 to 117 line 13.

36. For William's involvement in such a process see Thomson 1978c; also Thomson 1975:367-8, 385-6, 391. In this connection, it is interesting that the text of Suetonius in B is of the Z family, descended from a Norman parent (information from Professor R. H. Rouse).

37. This interest can be illustrated from Stubbs 1887:72-4, 125-6, 162, 164, 181-2, 185-7, 274-8.

38. He gives the Trojan origins of the Romans, after Dares and others, in Bodleian Library MS. Arch. Seld. B. 16, fos. 1-7v.

39. Southern 1970 and 1973; Gransden 1974. But note the comment on William's pan-European interest on p. 170.

40. Haskins 1927 has a chapter (8) entitled “Historical writing”, containing the challenging sentence “One of the best expressions of the intellectual revival of the twelfth century is to be seen in the writing of history”. Not many studies of the period have taken this further. Compare the remarkable statement of Southern 1971: “The intellectual climate of the twelfth century was not generally favourable to historical thought”. The writing of history is not discussed at all by Morris 1972, although he has a section entitled “The return to the past”. We clearly need a detailed study of twelfth-century historiography in relation to other contemporary intellectual and cultural phenomena.

Works Cited

Bautier, R.-H., and others. 1971. Odorannus de Sens, Opera omnia. Sources d'histoire médiévale, 4. Paris.

Bourassé, J.-J. 1861. Le cartulaire de Cormery précédé de l'histoire de l'abbaye et de la ville de Cormery. Mémoires de la Société Archéologique de Touraine. 12.

Brooke, Z. N. 1931. The English church and the papacy. Cambridge.

Ehlers, J. 1978. Die Historia Francorum Senonensis und der Aufstieg des Hauses Capet. Journal of Medieval History 4:1-25.

Fliche, A. 1909. Les sources de l'hisotriographie sénonaise au XIe siècle. Buletin de la Societé Archéologique de Sens 24: 149-206.

Gransden, A, Historical writing in England c. 550-c.1307. London.

Haskins, C. H. 1927. The renaissance of the twelfth century. Harvard.

Hlawitschka, E. 1965. Die Vorfahen Karls des Gorssen. In Karl der Grosse, ed. W. Braufels, 1:51-82. Düsseldorf.

Holder-Egger, O. 1890-1. Neues archiv16:602-6.

Hunt, R. W. and others. 1975. The survival of ancient literature. Bodleian Library. Oxford.

James, M. R. 1903. The ancient libraries of Canterbury and Dover. Cambridge.

Ker, N. R. 1969. Medieval manuscripts in British libraries. Oxford.

Kleinclausz, A. 1975. 1942. Eginhard. Paris.

Könsgen, E. 1975. Zwei unbekannte Briefe zu den Gesta regum Anglorum des Wilhelm von Malmesbury. Deustches Archiv 31: 201-14.

LEvinson, W. 1902. Zur Textgeschichte der VisionKaiser Karls II. Neues Archiv 27: 493-502.

Liebermann, F. 1879. Ungerdruckte Anglo-Normannische Geschichtsquellen.Strasbourg.

Martin, J. 1977. John of Salisbury's manuscripts of Frontinus and Gellius. Journal of the Warburg and Courtauld Institutes 40:1-26.

Poupardin, R. 1901. Le royaume de Provence sous les Carolingiens. Bibliothèque de l'Ecole des Hautes Etudes 131:324-32.

Prévost, A. Le and Delisle, L. 1855. Orderici Vitalis Historia ecclesiastica, 5. SHF. Paris.

Rose, V. 1893. Verzeichnis der lateinischen Handschriften der königlichen Bibliotek zu Berlin, Berlin.

Rottendorff, B. (ed.) 1638. Hugonis Floriacensis monachi Benedictini chronicon. Münster.

Rouse, R. H. and M. A. 1978. The medieval circulation of Cicero's “Posterior academics” and the De finibus bonorum et malorum. In Medieval scribes, manuscrits, and libraries; essays presented to N. R. Ker, ed. M. B. Parkes and A. G. Wartson, 351-2. London.

Somerville, R. 1972. The councils of Urban II: 1. Decreta Claromontensia. Annuarium historiae conciliorum, suppl. 1:88. Amsterdam.

Southern, R. W. 1970. The place of England in the twelfth-century renaissance. In Medieval humanism and other studies. Oxford.

Southern, R. W. 1973. Aspects of the European tradition of historical writing IV: The sense of the past. Transactions of the Royal Historical Society. 5th series 21:163.

Stevenson, W. 1959. Asser's life of King Alfred. Revised edition. Oxford.

Stubbs, W. (ed.) 1887 and 1889. William of Malmesbury, Gesta Regum Anglorum. 2 vols. R. S. London.

Thomson, R. M. 1975. The reading of William of Malmesbury. Revue bénédictine 85: 362-402.

Thomson, R. M. 1976a. William of Malmesbury, John of Salisbury and the Noctes atticae. Collection Latomus 145:367-89. Bruxelles.

Thomson, R. M. 1977. William of Malmesbury and the letters of Alcuin. Mediaevalia et humanistica, new series 8:147-61.

Thomson., R. M. 1978a. WIlliam of Malmesbury as historian and man of letters. Journal of ecclasiastical history 29:387-413.

Thomson, R. M. 1978b. William of Malmesbury's edition of the Liber pontificalis. Archivum historiae pontificiae 16:93-112.

Thomson, R. M. 1979. William of Malmesbury's reading; further additions and reflections. Revue beénédictine 89:318-9.

Wallace-Hadrill, J. M. 1962. The long-haired kings. London.

Whitelock, D. and others. 1961. The Anglo-Saxon chronicle. London and New Jersey.