APPENDIX B

The chronology of Duke William's campaigns between 1047 and 1054

The very difficult chronology of Duke William's wars during these years has been considered with great learning and with contrasted results by several continental scholars, such as L. Halphen,¹ R. Latouche,² and J. Dhondt,³ and I have been particularly indebted to the remarkable appendix which Henri Prentout added to his essay on the early life of Duke William, which was published posthumously in 1936.⁴ As will be seen, I have been able gratefully to accept, with only slight modifications, most of Prentout's conclusions on this matter, but the subject has been, and still is, so much a subject of controversy that it is necessary here to set out the evidence which has led me to adopt the dating of these events, which has been given in the preceding pages.

There is fortunately no need to question the accepted date of 1047 for the battle of Val-ès-Dunes. It must, however, be remarked that annals of Caen, and of Lire, give 1046, as does also Robert of Torigny in his supplement to the chronicle of Sigebert of Gembloux.⁵ It is not wholly impossible that this variation may perhaps have been due to an employment of the Lady Day reckoning for the beginning of the year which would here be relevant since the battle was evidently fought, as will be seen, before 25 March. Be that as it may, the date of 1047 is well attested. It is given, for instance, in the annals of Jumièges,⁶ of Saint-Évroul,⁷ and of Sainte-Colombe of Sens.⁸ Moreover, Ordericus Vitalis not only gives the year 1047 in his interpolations to William of Jumièges, but repeats this in his own history, where he relates the date convincingly to other events.⁹ The battle of Val-ès-Dunes may thus be assigned without undue hesitation to 1047, or more precisely (bearing in mind the normal practice of beginning the year at Christmas) tosome time after 25 December 1046; and it is in every degree probable that it occurred early in that period. The flooded waters of the Orne which played so large a part in the victory would be consistent with a date in winter or in very early spring.¹⁰

Although Henry I himself retired from Normandy after the battle, the war is known to have continued without interruption. Guy of Burgundy fortified himself at Brionne, and the siege of that castle began. Both William of Poitiers and William of Jumièges state that the defence was very stubborn,¹¹ but give no precise indication of its duration. Ordericus Vitalis on the other hand states that the siege lasted ‘for three years’ (per triennium),¹² and that this is not a mere verbal flourish is shown in another passage in hisHistory, where alluding to Duke Robert II (the Conqueror's son) he speaks of ‘Brionne … which his father with the aid of the French king could scarcely subdue in three years when Guy the son of Rainald of Burgundy defended it after the battle of Val-ès-Dunes.’¹³ It is of course possible to suggest that there might here be exaggeration,¹⁴ but to do so would be flying in the face of the evidence. There seems in fact no warranty for disregarding these curiously precise statements of an informed writer, and it appears impossible to avoid the conclusion that the siege of Brionne continued at least until the end of 1049.

These considerations have, moreover, some bearing on the most important and the most difficult question in this chronology, namely the date of William's campaigns during this period in Maine. A very strong tradition has assigned the warfare between the duke and Geoffrey Martel round Domfront and Alençon to 1048–1049, and this dating was endorsed in the brilliant reconstruction of this campaign made in 1906 by Louis Halphen in his notable book on the history of Anjou in the eleventh century.¹⁵ There it is suggested that the hostilities between the count and the duke began shortly after Geoffrey's attack on Château-du-Loir and the capture of Gervais, bishop of Le Mans, so that ‘before the autumn’ of 1048 Duke William and King Henry were together engaged in attacking Mouliherne in Anjou, and after October the duke on his own account captured Alençon and besieged Domfront, completing the operation in the early months of 1049. Though this was a very early book by this great scholar, published when its author was only twenty-six years of age, his detailed conclusions on this chronology have been accepted by nearly all subsequent writers on the subject. Nevertheless, before 1936 Henri Prentout had boldly challenged the accepted view, and placed the fighting round Alençon and Domfront in the winter of 1051–1052.¹⁶ Prentout's arguments (which do not appear to be widely known in England)¹⁷ seem to the present writer to be so compelling that they warrant a reconsideration of the matter.

It must immediately be emphasized how scanty is the positive evidence that can be adduced in favour of the traditional view. The warfare is not mentioned by the Angevin chroniclers, and it receives no notice in the early Norman annals. It is, therefore, necessary to depend in the first instance on the accounts given by William of Jumièges and William of Poitiers,¹⁸ and neither of these writers supplies a date. One difference between them may, moreover, be noted: William of Poitiers gives some prominence to a siege of Mouliherne by King Henry assisted by Norman troops, whilst William of Jumièges makes no mention of this episode. The two narratives, however, have this in common. They both place these events in their sequence between the capture of Brionne (which can hardly have fallen much if at all before the beginning of 1050), and the revolt of Count William of Arques which, as will be seen, began some time in the course of 1052. Whilst, therefore, the chronological arrangement of both William of Poitiers and William of Jumièges leaves very much to be desired, their testimony (for what it is worth chronologically) would seem to point to a date for these campaigns later than that which has traditionally been assigned to them.

Nor is the other evidence which has been cited in favour of the date of 1048–1049 for this war unequivocal.¹⁹ It is very reasonable to suggest that the capture of Gervaise at Château-du-Loir early in 1048 may have provoked hostilities between the king and the count of Anjou, but this does not necessarily imply that the duke of Normandy's expedition against Domfront occurred in that year. Finally, an allusion made by Anselm the monk in his description of the council of Rheims (1049) to hostilities then taking place between the king of France and his magnates, which has been held to refer to the war between Henry and Geoffrey, might with equal propriety be made to apply to almost any contemporary disturbance in the dominions of the king, and there was in fact in 1049 a siege of Neufchâtel-sur-Aisne near Rheims where the monk was writing.²⁰ On all grounds it would seem that the theory which place William's campaign round Alençon and Domfront in 1048–1069 is at best a plausible hypothesis which is inadequately supported by the surviving testimony.

This being so, it becomes imperative to consider whether some other interpretation will not better conform to the known facts. And in this connexion the blockade of Mouliherne falls for separate consideration, since scholars have perhaps been too hasty in treating it as inevitably part of the same question as that involving Domfront and Alençon.²¹ There is, in truth, nothing in William of Poitiers (the sole authority) to warrant this assumption. The arrangement of his treatise at this point merits in fact careful note.²²The passage respecting Mouliherne follows immediately after the account of the fall of Brionne. Then comes an account of the rising enmity of Geoffrey of Anjou towards Duke William. Then ensues a long passage about the accession of Edward the Confessor to the English throne, and not until after all this does there begin the description of the war round Alençon and Domfront. Moreover, Mouliherne is far removed from the scene of these latter campaigns: it is beyond Maine itself and in Anjou. The date of the blockade of Mouliherne must therefore be considered by itself. It may conceivably have occurred in 1048 for it is not far distant from Château-du-Loir. On the other hand, a letter from Geoffrey of Anjou to Pope Leo IX, the text of which has survived albeit in a late copy,²³suggests that the king's war against the count entered a new phase in 1051 and the Mouliherne blockade could thus be placed in the spring of that year. In that event the operations conducted by Duke William around Domfront (which occurred after the Mouliherne episode) would also fall inevitably in 1051. While, however, it is probable that the Mouliherne blockade did in fact take place in the spring of 1051, it is not impossible that it was further distinct in time from Duke William's later war in Maine.

Note must now be taken of certain dates in Angevin history which can be regarded as firmly established. These are:

Now, if these dates be considered it will appear that there is one time, and one time only, in this sequence when it might seem overwhelmingly probable that Geoffrey of Anjou, having at last established his position in the south of Maine, should direct his operations northward towards the Norman frontier, and that is in the period following his own establishment at Le Mans (March–April 1051). And there is, moreover, a considerable amount of converging testimony that this in fact occurred, provoking the retaliation of Duke William and the consequent war round the border fortresses. Thus Geoffrey's own letter to Leo IX, to which reference has already been made, specifically says that at this time Bishop Gervais, having obtained his release, broke faith with the count and went to Normandy where he urged the duke, and also the king, to take action in Maine. Duke William with the siege of Brionne now over, would at last be able, and doubtless ready, to respond. Indeed William of Poitiers says that he did so, strong in the knowledge that the domestic disturbances in Normandy were ended³¹ – a remark which would not have been appropriate had Brionne still be holding out.

Finally the same conclusion is imposed if the campaign round Domfront and Alençon is brought into relation with the subsequent revolt of Count William of Arques. One of the most significant events of the war in Maine was the sudden desertion of the duke by the count of Arques. At a critical moment in the siege of Domfront, we are told, he renounced his vassalage and departed, doubtless to organize his own rebellion in eastern Normandy.³² A remarkable chronological conformity thus seems to be here revealed, indicating afresh that the Domfront operation must surely have taken place in 1051–1052. For as late as the beginning of 1051, the count of Arques, still in possession of his dignity, was still in official harmony with the duke, having been then associated with him in the issue of grants for Saint-Wandrille,³³ so that his renunciation of vassalage can hardly have taken place before 1051. On the other hand, the count's desertion from Domfront during the winter of 1051–1052 would be the natural prelude to the count's own rebellion which, as will be seen, must have begun in the summer or autumn of the latter year.

The rebellion of Count William of Arques which was connected with the reconciliation between King Henry and Geoffrey Martel³⁴ took place between the capture of Domfront by Duke William and the battle of Mortemer. The battle of Mortemer can, moreover, be precisely dated. Neither William of Poitiers nor William of Jumièges supplies a date, but Ordericus Vitalis in interpolating the latter gives 1054³⁵ and in his own History he not only states that the battle occurred in 1054, but that it was fought in hieme ante quadragesimam.³⁶ Now, Ash Wednesday in 1054 fell on 16 February and Quadragesima Sunday on 20 February. The battle of Mortemer must thus be placed early in February 1054.

Two other events connected with the hostilities preceding the battle of Mortemer can also be dated with some precision. The reconciliation between Henry and Geoffrey Martel took place before 15 August 1052, since a dated charter shows them together at Orléans in amity on that day.³⁷ Secondly, the obituaries of Saint-Wulfram and Saint-Riquier unite in placing the death of Enguerrand, count of Ponthieu, on 25 October,³⁸ and this supplies also the date of the action at Saint-Aubin where Count Enguerrand was killed.³⁹ Now, as will be seen from the preceding argument, there are only two years to which this obituary could be assigned – 1052 and 1053. Of these 1052 is theoretically possible and should not be absolutely discarded. On the other hand, 1053 is much more probable, for the siege of Arques is known to have been very protracted, and the action at Saint-Aubin apparently came late in these operations. There are, it is true, discrepancies in the various accounts of this war. According to William of Poitiers,⁴⁰ Duke William originally came to Arques to place his own garrison there, and the fortress was later betrayed to the count; thereupon Duke William came again to Arques, fought an action in front of the castle, and left Walter Giffard to besiege it; and only after this did King Henry and the count of Ponthieu come to its relief with the consequent affray at Saint-Aubin. On the other hand, William of Jumièges omits any reference to the first betrayal of the castle, and Ordericus Vitalis speaks only of one march by Duke William into Talou.⁴¹ But whatever version of these events be accepted, it will be seen that before the eventual fall of Arques there occurred at least one and possibly two expeditions by Duke William, not to mention the engagement which involved the death of Hugh of Morimont, and the action at Saint-Aubin which cannot have occurred later than 25 October 1053. It seems, therefore, impossible to escape the conclusion that the rebellion of William of Arques must have begun in 1052, and since the rebels appealed to King Henry⁴² it should be related to the reconciliation between the French king and Count Geoffrey which was complete by 15 August of that year.

After a review of the evidence, and with all possible reserve in view of its difficult nature, I have therefore adopted the following chronology for these events. The battle of Val-ès-Dunes took place very early in 1047, and thereafter ensued the siege of Brionne which lasted until the end of 1049. In the meantime Geoffrey of Anjou (late 1047 or early 1048) had attacked Château-du-Loir and captured Bishop Gervais. It is possible, though unlikely, that the king with Norman assistance blockaded Mouliherne about this time, but that event should more probably be assigned to the spring of 1051. In March–April 1051 Geoffrey Martel occupied Le Mans, and thereafter moved up to the Norman frontier. Duke William, urged on by Bishop Gervais, who was now in Normandy after his release, thereupon in the autumn of 1051 came to the relief of Domfront and Alençon. The siege of Domfront lasted through the winter months, and was concluded early in 1052. During the siege Count William of Arques renounced his vassalage, went eastward to prepare his own rebellion. He appealed to the king of France, who became reconciled to Geoffrey Martel before 15 August 1052. The war involving the siege of Arques began in the summer or autumn of 1052, and was still continuing on 25 October 1053. The castle fell very late in 1053, and early in the next year King Henry launched his double invasion of Normandy which was repelled at the battle of Mortemer in February 1054.

1 Comté d'Anjou (1906), pp. 70–80.

2 Comté du Maine (1910), pp. 27–32.

3 ‘Les Relations entre la France et la Normandie’ (Normannia, vol. XII (1939) pp. 465–486); ‘Henri I^er, l'Empire et l'Anjou’ (Revue belge de Philologie et d'Histoire, vol. XXV (1946), pp. 87–109).

4 Histoire de Guillaume le Conquérant – Le Duc De Normandie (Mém. Acad. Nat. de Caen, vol. VIII, 1936). This is an incomplete work issued posthumously.

5 Rec. Hist. Franc., vol. XI, pp. 166, 366.

6 Ed. J. Laporte, p. 55.

7 Ord. Vit., vol. V, p. 157.

8 Rec. Hist. Franc., vol. XI, p. 292.

9 Will. Jum., p. 171; Ord. Vit., vol. I, p. 182; vol. II, p. 373; vol. III, p. 159.

10 Above, p. 50.

11 Will. Poit., pp. 19–21; Will. Jum., p. 123.

12 Ord. Vit., vol. IV, p. 335.

13 Ibid., vol. IV, p. 335.

14 Cf. Freeman, Norman Conquest, vol. II, p. 262; Stenton, William the Conqueror, p. 85.

15 Op. cit., pp. 70–73.

16 Op. cit., pp. 140–144.

17 I note, however, that de Bouard places the action round Domfront, without further comment in 1050–1051 (Guillaume le Conquérant, p. 41).

18 Will. Jum., pp. 125–127; Will. Poit., pp. 23, 37–39.

19 Cf. Prentout, loc. cit.

20 Ibid. Cf. Rec. Hist. Franc., vol. XI, p. 465.

21 I am not wholly convinced by Prentout's argument in this sense (ibid., p. 142).

22 Will. Poit., pp. 22–40.

23 Sudendorf, Berengarius Turonensis (1851), Appendix VIII.

24 Actus Pont. Cenomm.; Halphen, op. cit., p. 71; Latouche, op. cit., p. 28.

25 Above, p. 58.

26 Ann. S. Maxence (Marchegay et Mabille, Chroniques des Églises, p. 398).

27 Necrologie de la Cathédrale du Mans (ed. Busson and Ledru), p. 72.

28 Actus Pont. Cenomm.

29 Ibid., and Sudendorf, loc. cit.

30 Rec. Hist. Franc., vol. XI, p. 590; Soéhnée, Cat. Actes, Henri I, no. 91.

31 Will. Poit., p. 65.

32 Will. Poit., p. 63.

33 Chevreux et Vernier, Archives, plate IV.

34 Ord. Vit., vol. I, p. 184.

35 Will. Jum., p. 180.

36 Ord. Vit., vol. III, pp. 160, 237.

37 Soéhnée, op. cit., no. 91.

38 C. Brunel, Rec. Actes des Comtes de Ponthieu (1930), p. iv.

39 Will. Jum., p. 120.

40 Will. Poit., pp. 55–61.

41 Will. Jum., p. 119; Ord. Vit., vol. III, pp. 42, 232, 233.

42 Ibid., vol. I, p. 184.