Research Notes:
Complete Peerage provides the following biography of Roger I de Montgomery:[1]
Roger I de Montgomery, seigneur of Montgomery and vicomte of the Hiesmois, witnessed in 1031 or 1032 a charter of Robert I, Duke of Normandy, for the abbey of St. Wandrille. Between 1028 and 1035 he restored to the abbey of Jumièges a market at Vimoutiers, near Montgomery, of which he had deprived the monks. During the troubles of Duke William's minority he was exiled and went to Paris on account of the murder of Osbern, s. of Herfast, the steward, by his s. William de Montgomery. At some date unknown he founded a collegiate church for twelve secular canons at Troarn. He m. Josceline, a niece of the duchess Gunnor. The date of his death is uncertain.
Complete Peerage provides the following evidence concerning the wife of Roger I de Montgomery: [2]
Robert de Torigny in his continuation of William de Jumieges (bk. viii, c. 35—ed. Marx, p. 321) makes the mother of Roger II de Montgomery to be Josceline, da. of Wevie, a sister of Gunnor; in c. 37 he states Wevie's husband to have been Osbern de Bolbec. He calls Josceline's husband Hugh de Montgomery, but the father of Roger II was undoubtedly named Roger. The mistakes in these genealogical chapters, 35-37, as to which see G. H. White's article on "The Sisters and Nieces of Gunnor, Duchess of Normandy " in Genealogist, N.S., vol. xxxvii, pp. 57-65, 1 28-1 32, are too numerous to make it possible to trust them in default of corroboration, but the statement as to Josceline is in part corroborated by Ives, bishop of Chartres. In a letter of the year 1 1 14 to Henry I (Rec. des Hist, de France, vol. xv, p. 167) he points out that a projected marriage between a natural daughter of the King and Hugh, son of Gervase de Chateauneuf, would be invalid on the ground of consanguinity. He traces Hugh's pedigree as follows: Senfrie, sister of Gunnor, had a daughter Josceline, the mother of Roger de Montgomery, who had a daughter Mabel married to Gervase de Chateauneuf, by whom she was the mother of Hugh. It will be observed that this account makes Josceline the daughter of Senfrie or Sainfrie, who according to Robert de Torigny (loc. cit.) was Gunnor's sister; the name of Josceline's husband is not given. The bishop seems to be writing with due sense of responsibility, for the letter contains this remarkable passage: "Quod enim dicimus, non ex conjecturis facimus, quia prae manibus habemus scriptam genealogiam, quam scribi fecerunt nobiles viri de eadem tribu progeniti, et parati sunt ante judices ecclesiasticos eandem genealogiam in tuto loco computare et secundum legum instituta probare." Moreover the protest was successful, since Hugh did not marry the King's daughter but a daughter of Waleran, Count of Meulan. If however another of these genealogies of Robert de Torigny be accepted, Hugh nevertheless married within the prohibited degrees, since his bride descended from another of Gunnor's sisters (see ante, vol. vii, p. 521, note "c"). Taken as a whole the evidence seems to show that Roger I's wife was Josceline, a niece of Gunnor.
Kathleen Thompson suggests that [3]
It is possible, however, to construct a more plausible genealogy for the Montgomery family without entirely rejecting the work of Robert of Torigni. Torigni states that Joscelina married a Hugh of Montgomery, while Roger II's own charter to Troarn gives his father's name as Roger. This error may not invalidate the genealogy, but rather imply that a generation has been omitted in transmission. It may be that Joscelina niece of the duchess married, probably in the nine-nineties, a Hugh from Montgommeri, for the name Hugh does recur in the Montgomery family, and that her son, Roger I, was born about the year 1000. [Footnote: The name Hugh was used for Roger II's elder brother, nephew and second son (see Table I). Roger II was later to give Riville (Manche. cant. Quettehou) to Troam (Sauvage, Troam. p. 349). Since Arfast, brother of the duchess Gunnor, held lands in this area of the Cotentin, it is possible that this outlier came into Montgomery possession through Joscelina's marriage (Douglas, 'Ancestors'. p. 68).] This elevation of Joscelina by one generation is more satisfactory chronologically and helps to explain Roger II's otherwise puzzling failure to re-use her name for one of his daughters. Ivo of Chartres's genealogy which gives only the line of descent and not the spouses of the individuals mentioned does not contradict this theory, only suggests that the blurring of the genealogy had already taken place in the eleventh century.
[1] George E. Cokayne and Geoffrey H White, ed., The Complete Peerage, rev., Vol. 11, Rickerton to Sisonby (London: St Catherine Press, 1949), 382-83.
[2] George E. Cokayne and Geoffrey H White, ed., The Complete Peerage, rev., Vol. 11, Rickerton to Sisonby (London: St Catherine Press, 1949), 383, footnote c.
[3] Kathleen Thompson, "The Norman Aristocracy before 1066: the Example of the Montgamerys," Historical Research 40 (1987), 251-263 at 254 and 258.