APPENDIX C
Few episodes in the Conqueror's life have given rise to more controversy than his marriage to Matilda, daughter of Baldwin V, count of Flanders, by Adela, daughter of Robert I, king of France. Even the precise date of the marriage is uncertain. It was projected in or before 1049, but it had not taken place by the autumn of that year when it was forbidden by Leo IX at the time of the council of Rheims.1 On the other hand, it had been celebrated before the end of 1053, in which year Matilda appears as the duke's consort in a dated charter given to Holy Trinity, Rouen.2 Within these dating limits it is, however, very hard to particularize. The annals of Tours, themselves an unreliable compilation in this matter, were erroneously cited by Freeman in favour of 1053,3 but the date itself is not impossible, for the imprisonment of Leo IX by the Normans after the battle of Civitate (June 1053)4 might have provided an occasion for defying the papal prohibition. On the other hand, there are reasons for placing the marriage earlier, and for viewing it in connexion with the transformation in the relations between the duke of Normandy and the king of France which took place in 1051–1052. Thus Mlle Foreville gives 1050–1051 for the marriage,5 and Professor de Bouard places it sans doute en 1050 ou peu après.6These opinions carry conviction, and some of the facts in the early career of Robert, William's eldest son, can be cited in its support. Robert was Matilda's eldest child, and he was born in wedlock. His birth has been placed circa 1054, but it is not impossible that he was born earlier.7 He is described as having been adolescens in 1066,8 and in that year also he was cited as confirming a charter for Marmoutier quia scilicet maioris iam Me aetatis ad praebendum spontaneum auctoramentum idoneus esset.9Although therefore the tender age of Matilda (see below) must be taken into account in assigning a date for her first pregnancy, it seems not unreasonable to place her marriage in 1050–1051,10 but definite proof of this is lacking.
The view once held that Matilda was already married when William sought her hand, and was then the mother of a daughter, Gundrada, later the wife of William of Warenne, has now been conclusively disproved by the researches of Chester Waters11 and Sir Charles Clay.12 There is no reason to suppose that Gundrada was the daughter either of William or Matilda. Moreover, it is unlikely that Matilda herself was of an age to be a mother in 1049 when the papal ban was pronounced. Her age cannot be precisely ascertained but the marriage of Baldwin V to her mother, Adela, was apparently not consummated until 1031, for William of Jumièges states that this was one of the factors in stimulating Baldwin V's rebellion against his father in that year.13 On this showing Matilda could not have been more than seventeen in 1049, and she may have been younger since there is no evidence to show that she was the eldest of the four children given by Adela to Baldwin V.
All theories respecting the reasons of the papal prohibition of the match between William and Matilda are therefore now based on the suggestion that (as hinted by Ordericus Vitalis)14 the parties were within the prohibited degrees. Scholars, however, have differed sharply as to the nature of the alleged relationship between them. Three views on the matter may be briefly noted:
1. It has been suggested15 that the prohibition was due to a marriage between Duke Richard III of Normandy and Matilda's mother, Adela of France. There are, however, objections. Certainly, Richard III married a woman named Adela, but it is doubtful if this was Adela of France, and, in any case, a marriage between Richard III and Adela of France could not have been consummated.16
2. It has been suggested17 that the ban was based on a common descent of William and Matilda from Rolf the Viking, which would have made them cousins in the fifth degree. Whether this descent (which is itself not beyond question) would have been sufficient for the prohibition even if it had been known in Rome is perhaps somewhat doubtful.
3. It has been held on the authority of William of Jumièges18 that Baldwin IV of Flanders married not only Ogiva, daughter of Richard, duke of the Ardennes, who was the mother of Baldwin V, but also a daughter of Richard II of Normandy. This marriage, if it occurred, might perhaps have been made to serve as a reason for the ban.
All these theories are open to some criticism, and despite the erudition that has been lavished upon them, the question cannot be regarded as settled.
The marriage of William and Matilda was by all accounts very happy, and it was certainly fruitful. Matilda bore her husband four sons and at least five daughters.19 The sons were Robert, later duke of Normandy; Richard; and William and Henry, subsequently kings of England. Richard was accidentally killed in the New Forest at an early age.20 The date of the mishap is not known, but since he was young at the time, and since he was Matilda's second son (though not necessarily her second child), his death can reasonably be placed between 1070 and 1080, and perhaps circa 1075.
The question of the daughters of William and Matilda is more complicated, and the essential evidence may be briefly tabulated as follows:
A. William of Poitiers states:21 (i) that a daughter of William (unnamed) was betrothed to Herbert, count of Maine; (ii) that a daughter of William (unnamed) was sought for in marriage by two rival kings of Spain who were brothers, one of whom has been reasonably identified as Alphonso IV, later king of Leon; and (in) (by implication) that a daughter of William (unnamed) was at one time betrothed to Harold of Wessex.
B. Ordericus Vitalis22 mentions five daughters of William's marriage whom he names and describes as follows: (i) Agatha, who was betrothed successively to Harold of Wessex and Alphonso of Spain. She protested vigorously against going to Spain, and died a virgin, being buried at Bayeux; (ii) Adeliza, who undertook religious vows early in life and lived under the protection of Roger of Beaumont; (iii) Constance, who married Alan IV of Brittany; (iv) Adela, who married Stephen I, count of Blois; and (v) Cecily, who became abbess of Holy Trinity, Caen.
C. William of Malmesbury23 mentions five daughters, namely Cecily, Constance, and Adela as above, and two other daughters (unnamed), one of whom, he says, was betrothed to Harold, and the other to Alphonso.
D. Robert of Torigny24 at a later date speaks of four daughters, Cecily, Constance, and Adela as above, and also ‘Adeliza’, who according to him was betrothed to Harold.
E. Domesday Book25 mentions a daughter of William named Matilda.
Both the similarities and the contradictions in these accounts are interesting. Cecily, Constance, and Adela are well known from their subsequent careers. Agatha and Adeliza have on the evidence been thought to represent only one person.26 It may be so, but I am personally inclined here to accept Ordericus's curiously precise statement about Adeliza at its face value. The remark of William of Poitiers at least helps to confirm his statement that one of these daughters was betrothed to Alphonso of Spain, though whether one of them was betrothed to Herbert of Maine as well as to Harold of Wessex must remain in doubt. It will be noted further that none of these chroniclers mentions Matilda, and it might be tempting to doubt her existence or her legitimacy were it not that a Caen narrative apparently mentions her in connexion with both her mother and her sister Cecily, thus confirming the reference in Domesday Book.27
It may therefore be concluded that William and Matilda had four sons, born in the following order:
1. Robert, later duke of Normandy. Born 1051–1054. Died 10 February 1134.
2. Richard. Born before 1056. Died circa 1075?
3. William, later king of England. Born 1056–1060. Died 2 August 1100.
4. Henry, later king of England. Born late 1068. Died 1 December 1135
It would appear also that William and Matilda may have had six daughters who (without reference to seniority) might be enumerated as follows:
1. Agatha, betrothed successively to Harold, earl of Wessex, and to Alphonse of Leon (and possibly previously to Herbert, count of Maine). Died a virgin.
2. Adeliza.
3. Cecily, born before 1066, subsequently abbess of Holy Trinity, Caen. Died 1127.
4. Adela, married, 1080, Stephen I, count of Blois. Died 1137.
5. Constance, married, 1086, Alan IV, count of Brittany. Died 1090.
6. Matilda.
Dogmatism would here, however, be out of place. The separate existence of Agatha and Adeliza is not certain, and the evidence about Matilda is less than satisfactory.28 The relative ages of the daughters is moreover not known, and one at least of them, Cecily, was born before Henry I. Finally, it may deserve a note of surprise (and admiration) that a lady of such diminutive size as was William's wife29 should have produced so large a family before her own death in 1083.
1 Hefele-Leclerc, Histoire des Conciles, vol. IV, part II, p. 1018.
2 Cart. S. Trin. Roth, no. XXXVII.
3 Freeman, Norman Conquest, vol. IV, note O.
4 Chalandon, Domination normande, vol. I, p. 137.
5 Ed. Will. Poit., pp. 46, 66.
6 Guillaume le Conquérant, p. 36.
7 David, Robert Curthose, p. 5.
8 Will. Malms., Gesta Regum, p. 450.
9 Bertrand de Broussillon, Maison de Laval, vol. I, p. 45, no. 30.
10 Handbook of British Chronology (R. Hist. Soc. (1961), p. 31).
11 Gundrada de Warenne (Exeter, 1884).
12 Early Yorkshire Charters, vol. VIII, pp. 40–46.
13 Will. Jum., pp. 103–104.
14 Interp. Will. Jum., pp. 181–182.
15 W. H. Hutton, in Dict. Nat. Biog., sub. ‘Mathilda’.
16 Adela of France was brought to the Flemish court when the girl was in cunis, and she was kept there for some years – annos usque ad nubiles – until the marriage could be consummated, an event which apparently occurred in 1031 (Will. Jum., p. 103). But Richard III was duke from 1026 to 1028 and on this reckoning Adela of France could then have been little more than an infant.
17 Prentout, ‘Le marriage de Guillaume le Conquérant’ (Études sur quelques points d'histoire de Guillaume le Conquérant, Caen, 1930) – a most valuable article which discusses much of the evidence here considered.
18 Will. Jum., p. 88.
19 Freeman, Norman Conquest, vol. IV, note O; Handbook of British Chronology, loc. cit.
20 Will. Malms., Gesta Regum, p. 332.
21 Will. Poit., pp. 89, 143, 230.
22 Ord. Vit., vol. II, pp. 189, 391, 392; vol. III, p. 159.
23 Gesta Regum, p. 333.
24 Interp. Will. Jum., pp. 317, 318.
25 D.B., vol. I, fol. 49.
26 Handbook of British Chronology, loc. cit.
27 Freeman, Norman Conquest, loc. cit.
28 The matter is complicated by the fact that Ordericus (vol. II, p. 182) says that one of William's daughters was betrothed to Edwin of Mercia.
29 Above, pp. 369, 370.