APPENDIX E

The chronology of King William's campaigns between 1073 and 1081

The chronology of the Conqueror's campaigns between 1073 and 1081 is beset with difficulty, and it has engaged the attention of many scholars as distinguished as Freeman, Miss Norgate, Louis Halphen, Sir Frank Stenton, M. Latouche, A. Fliche, and Professor C. W. David. Since, however, among these scholars there has been considerable disagreement, the matter may properly be ventilated afresh. Though in debt to all these authorities in what follows, I am myself inclined to accept, although with some modifications, the solution propounded in 1906 by Halphen,¹ and developed later by Fliche² and Professor David.³ The matter is, however, sufficiently complicated to warrant the evidence being restated.

An essential part of the problem concerns the question of how far, if at all, the statements of Ordericus Vitalis on this matter can be accepted.

Apart from two detached notices concerning the revolt of Robert and the battle of Gerberoi,⁴ the narrative of Ordericus of these campaigns is contained in his History in Book IV, chapters XII, XIII, XIV, and XVII.⁵ Chapters XV and XVI contain a long digression inspired by the notice of the death of Earl Waltheof at the end of chapter XIV. The whole thus forms a unity, and begins shortly after Ordericus has lost the support in his narrative of the lost portion of the history of William of Poitiers.

The sequence of events given by Ordericus is as follows:

1. William's campaign in Maine. Entry into Le Mans.

2. Fulk le Rechin attacks La Flèche. (According to Ordericus, Fulk is supported by ‘Count Hoel’ with a large force of Bretons. William leads an army of Normans and English against him. A great battle is averted by the intervention of clerics.)

3. A pact is arranged between William and the count of Anjou. (According to Ordericus this was concluded at a place called Blancalanda vel Brueria, and the writer adds that this peace remained unbroken until the death of the Conqueror.)

4. The revolt of the earls in England and the death of Waltheof.

5. William's campaign in Brittany, culminating in his unsuccessful siege of Dol. (According to Ordericus, Dol was relieved by Alan Fergant, ‘count of Brittany’. William was compelled to retreat and Alan Fergant married Constance, William's daughter.)

William's campaign in Maine (O. V. 1) can be precisely dated: it occurred in 1073 and probably before 30 March of that year.⁶ Similarly, there is no question that the revolt of the earls (O. V. 4) took place in 1075.⁷ These notices by Ordericus thus need no further comment in this place. The remainder, however, call for careful examination and criticism.

In the first place, the campaign of William in Brittany needs consideration. The Anglo-Saxon, Breton, and Angevin annals, together with the charters of Philip I, make it quite certain that this campaign and William's defeat at Dol took place in September – early November 1076:⁸ the campaign followed indeed logically upon the revolt of Earl Ralph in England, and it was connected also with Count Hoel's warfare with the rebel magnates of Brittany.⁹ This being so, the narrative of Ordericus at this place becomes at once unintelligible for the following reasons:

(i) He makes no mention of King Philip's intervention at Dol, which certainly took place and was probably decisive.

(ii) Count Hoel did not die until 1084,¹⁰ and it was then that he was succeeded by his son Alan Fergant. Unless Ordericus has here given Alan the comital style simply as a tribute to birth, his bringing him into the story in this manner, in 1076, would seem to be an error.

(iii) It is in the highest degree unlikely that Constance married Alan Fergant in 1076 or 1077. Ordericus, himself, seems elsewhere to imply that she was still unmarried in 1081, and there is positive evidence to suggest that her marriage to Alan Fergant took placein 1086.¹¹ It may be further noted that Ordericus is not, in general, very sure about Alan Fergant, since elsewhere he calls him count of Nantes, whereas he was the son of Hoel, count of Cornouailles.¹²

(iv) If, as the Breton annals assert, Hoel acted in alliance with William at the time of the siege of Dol, it would be very unlikely (though not impossible) that Alan Fergant should have acted against his father in this campaign, and this improbability is enhanced by the fact that in 1077 Alan rescued his father from the Breton rebels.¹³

Some of these difficulties were well appreciated by Freeman in a remarkable appendix,¹⁴ whose value is only partially vitiated by a misconception of the dating of one of the Angevin annals.¹⁵ His guess that Alan Fergant may have been betrothed to Constance in 1077, and only married to her nine years later,¹⁶ would not, however (even if it were plausible), bring Ordericus's account into line with the other sources. Indeed, Freeman himself hazarded the suggestion (which had before been made by Lobineau)¹⁷ that Ordericus was here referring to some warfare in Brittany at a later date. Either this was so – and there is no other evidence to support it – or Ordericus was here relying on erroneous information. In any case, his account adds nothing of value to our knowledge of the campaign in Brittany in 1076. It can be set aside.

Further difficulties arise in connexion with the account given by Ordericus of the attack on La Flèche by Fulk. As will be seen, Ordericus in his narrative places this after the campaign in Maine and the fall of Le Mans.¹⁸ For this reason, Freeman, Stenton, and (with some hesitation) Miss Norgate place the attack on La Flèche in 1073.¹⁹ It would seem, however, that Ordericus himself does not give warranty for this conclusion. After speaking of the fall of Le Mans, he suggests that an interval may have occurred before the subjugation of Maine,²⁰ and in any case the affair at La Flèche (as described by Ordericus) is made by that writer to lead directly up to the pact between William and Fulk at Blancalanda vel Brueria.²¹ Thus, even if the narrative of Ordericus was here to be taken at its face value, it would be reasonable to place the episode he describes as taking place at La Flèche at any time between the fall of Le Mans (spring, 1073) and the pact at Blancalanda vel Brueria for which he supplies no date.

In face of such obscurity it is necessary to turn to other evidence.

The Angevin annals make it clear that there were two attacks about this time by Fulk Rechin on La Flèche. The first (which was unsuccessful) occurred in 1076, 1077, or possibly 1078.²² The second (which was successful and involved the burning of the castle) took place in 1081 – that is to say after the revolt of Robert, and William's reverse at Gerberoi.²³ It would seem, therefore, that Ordericus's account must refer to one or other of these, unless (as is very probable) he has confused the two. It will be recalled, moreover, that he places his notice immediately before, and in connexion with, the pact between William and Fulk at Blancalanda vel Brueria. The date of the pact thus described, therefore, needs to be settled, and here the description given by Ordericus of Fulk's army during the attack on La Flèche is relevant. He remarks that Fulk on this occasion was assisted by ‘Count Hoel’ and ‘a very large force of Bretons’.²⁴ Now, it is likely enough that Breton troops should have supported their Angevin allies in Brittany in following up their success at Dol, but these would have been from among the Breton rebels who in 1076 had been fighting not only against William but against Hoel.²⁵ It seems, in short, inconceivable that Hoel should have attacked La Flèche and the Normans in Maine in conjunction with Fulk at any time during 1076 or 1077: indeed, during the latter year he himself fell a prisoner to the Breton rebels and was only rescued by his son Alan Fergant.²⁶ Two alternative conclusions seem thus to be imposed. Either Hoel must be eliminated from Ordericus's story, or the treaty which he describes between Fulk and William must be connected with the second attack on La Flèche – that is to say, in 1081. Consequently it is important that the annals ‘of Renaud’ specifically speak of a pact between William and Fulk in 1081, and in terms which appear to supplement Ordericus's description.²⁷ Not only, as Ordericus says, was there a general amnesty, and the performance of Robert's homage to Fulk ut minor majori, but, as the annals ‘of Renaud’ would add, William had to offer hostages to Fulk.²⁸ Finally, the remark of Ordericus that this pact gave peace between Anjou and Normandy until the death of the Conqueror,²⁹ though in no case fully accurate, would be understandable in relation to a treaty ratified in 1081, but quite incomprehensible in connexion with a treaty made in 1076–1078 when Robert's rebellion, and the battle at Gerberoi, and the second attack on La Flèche were still to come. On all grounds it seems impossible (or at least very hazardous) to place the so-called ‘treaty of Blanchelande’ elsewhere than in 1081.

What, then, happened between William and Fulk in 1077–1078? It is certain that in 1077 there occurred a treaty between William and King Philip ‘which did not last long’,³⁰ and it would seem unlikely that the French king (so soon after the events in Brittany in the autumn of 1076) would here have acted without associating his Angevin vassal and ally in any arrangements which were then made. It is thus improbable that Fulk's first attack on La Flèche occurred after the pact between William and Philip, and equally unlikely that it took place before the end of the Breton campaign in early November 1076. The Angevin annals, it will be recalled, give three dates for this attack – 1076, 1077, and 1078. The last year is given in the annals of Saint-Florent of Saumur, but these are often a year in error,³¹ and for the reasons given above I am inclined to reject it.³² MS. ‘D’ of the annals of Saint-Aubin gives 1076, but since the numeral occurs twice in the manuscript, it has the appearance of a repeated entry, and MSS. ‘B’, ‘C’, and ‘A’ of these annals give 1077, which would also be more in keeping with the circumstantial evidence.³³ I, therefore, place the first attack of Fulk on La Flèche early in 1077 or perhaps very late in 1076.³⁴

This unsuccessful operation by Fulk might well have resulted in some temporary agreement between him and William, made in conjunction with the pact entered into between William and King Philip in 1077. It is therefore significant that a charter for Saint-Vincent-du-Mans given about this time is dated by reference to a pact being made between King William and Count Fulk at a place called Castellum Vallium.³⁵ This charter certainly passed before 5 November 1080, for it is witnessed by William as abbot of Saint-Vincent, and this William was at that date appointed bishop of Durham. Consequently, if the so-called ‘treaty of Blanchelande’ was made, as has been here suggested, in 1081, this truce must be distinct from it. It would thus seem reasonable to place it in 1077–1078, and in direct relation to the pact of 1077 made between King William and Philip.

There only remain to be considered the events between 1077 and 1081. Malcolm's invasion of northern England is fixed by the Anglo-Saxon Chronicle as having occurred between 15 August and 8 September 1079.³⁶ Again, Robert's counter-raid on Scotland is firmly placed by ‘Simeon of Durham’ in 1080, and it must have occurred after 14 July of that year when Robert and his father were still in Normandy.³⁷ The dates of Robert's first rebellion against his father, and of the battle of Gerberoi present, however, some difficulties. Freeman assigned the siege of Gerberoi to the winter of 1079–1080;³⁸ this, however, is demonstrably wrong. Ordericus Vitalis states that the siege (which lasted some time) began immediately after Christmas, and the Anglo-Saxon Chronicle (which normally begins the year at Christmas) places it in 1079.³⁹ The significance of this combined testimony is, moreover, conclusively confirmed by a charter of King Philip I of France⁴⁰ which is witnessed by Robert the steward, and which not only refers to the siege of Gerberoi as in progress, but contains a dating clause⁴¹ which could indicate Christmas 1078, but could not possibly indicate Christmas 1079 – indeed during the course of 1079 Robert was to be replaced as steward by Adam.⁴² The siege of Gerberoi thus took place in the last days of 1078 and was concluded during the first weeks of January 1079.

Robert's first quarrel with his father obviously broke out some time before this, but it occurred later than 13 September 1077, when Robert took part with his father at the dedication of Saint-Stephen's, Caen.⁴³ His rebellion must, therefore, have begun either very late in 1077, or, more probably, early in 1078. After Gerberoi, he was in due course reconciled with William, and, certainly, this reconciliation took place before 12 April 1080, when he is to be found once again at his father's court.⁴⁴

The chronology of these events is so obscure, and so many points respecting it still await full elucidation, that it would be signally rash to dogmatize too confidently about it. These remarks will, however, explain why, with whatever hesitations, I have adopted the following sequence in this book:

1 Comté d'Anjou, esp. pp. 181–183.

2 Philippe I, pp. 271–274.

3 Robert Curthose, chaps. I and II.

4 Ord. Vit., vol. II, pp. 294–298, 386 387.

5 Ibid., vol. II, pp. 254–267, 290–292.

6 AS. Chron., ‘E’, s.a. 1073; Regesta, vol. I, nos. 67, 68.

7 AS. Chron., ‘E’, s.a. 1075.

8 AS. Chron., ‘E’, s.a. 1076; Chron. Britannicum and Chron. S. Brieuc (Rec. Hist. Franc., vol. XI, p. 413; vol. XII, p. 566); Annales Saint-Aubin and Annales ‘de Renaud’ (printed in Halphen, Annales Angevines, pp. 5, 88); Prou, Rec. Actes – Philippe I, no. LXXXII.

9 Ann. S. Brieuc (Rec. Hist. Franc., vol. XI, p. 413).

10 Clay, Early Yorkshire Charters, vol. IV, p. 84.

11 Ord. Vit., vol. III, p. 28; Rec. Hist. Franc., vol. XII, pp. 559, 562, 563.

12 Ord. Vit., vol. II, p. 392; Clay, loc. cit.

13 Rec. Hist. Franc., vol. XI, p. 413; vol. XII, p. 566; Morice, Hist. de Bretagne, Preuvies, vol. I, p. 102; Chron. Quimperlé (Rec. Hist. Franc., vol. XII, p. 561).

14 Norman Conquest, vol. IV, p. 817.

15 As shown by Marchegay and Mabille (Chroniques des Églises, p. 12) the entry MDLXXXVI in the Annales ‘de Renaud’ is a scribal error for MDLXXVI.

16 Freeman, loc. cit.

17 Histoire de Bretagne, vol. I, p. 104.

18 Ord. Vit., vol. II, p. 255.

19 Freeman, Norman Conquest, vol. IV, p. 561; Stenton, William the Conqueror, p. 313; K. Norgate, Angevin Kings, vol. I, p. 257.

20 Ord. Vit., vol. II, p. 255.

21 Ord. Vit., vol. II, p. 258.

22 Halphen, Comté d'Anjou, p. 311, no. 233; Halphen, Annales, pp. 5, 119; David, Robert Curthose, p. 32, note 71.

23 Halphen, Annales, pp. 5, 88.

24 Ord. Vit., vol. II, p. 256.

25 Above, pp. 233–235.

26 Rec. Hist. Franc., vol. XII, p. 561.

27 Halphen, Annales, p. 88; Comté d'Anjou, pp. 183, 184.

28 Halphen, Annales, p. 88; Ord. Vit., vol. II, p. 257.

29 Ord. Vit., vol. II, p. 258.

30 AS. Chron., ‘E’, s.a. 1077.

31 Halphen, Annales, p. 119.

32 ‘But see David, op. cit., p. 32, note 71.

33 Halphen, Annales, p. 5.

34 Contrast Halphen, Comté d'Anjou, p. 182.

35 Cart. S. Vincent du Mans, no. 99. This passed ‘eo tempore quo Willelmus rex Anglorum cum Fulcone Andavagensi comite juxta castellum Vallium trevisam accepit’.

36 AS. Chron., ‘E’, s.a. 1079: ‘between the two feasts of Saint Mary’.

37 Opera, vol. II, p. 211; Gall. Christ., vol. XI; Instrumenta, col. 266.

38 Op. cit., pp. 646, 647.

39 Ord. Vit., vol. II, p. 387; AS. Chron., ‘E’, s.a. 1079.

40 Prou, Rec. Actes – Philippe I, no. XCIV.

41 Actum publice in obsidione predictorum regum videlicet Philippi regis Francorum et Guillelmi Anglorum regis circa Gerborredum, anno Incarnati Verbi, millesimo septuagesimo VIII, anno vero regni Philippo regis Francorum XIX.

42 Prou, op. cit., no. XCVII.

43 Gall. Christ. vol. XI; Instrumenta, col. 72; Regesta vol. I, no. 96.

44 Cart. S. Trin. Roth., no. LXXXII.