The Henry Project: The Ancestors of King Henry II of England

Judith of Flanders

(Wife of Tostig of Northumbria and Welf of Bavaria)

Date of Birth: 1030×1036.
(based on the fact that her father's first wife died in 1030, and her father died in May of 1035)
Place of Birth: Unknown.

Date of Death: 5 March 1094.
[Decker-Hauff (1956), 46 (n. 34), citing the Weingarten Necrology]
Place of Death: Unknown.

Father: Baldwin IV, d. 1035, count of Flanders.

Mother: NN of Normandy, daughter of Richard II, duke of Normandy.

The primary sources are contradictory regarding the parentage of Judith, but the solution given above may be regarded as reasonably certain. The evidence is discussed in detail in the Commentary section below.

Spouses:

(1) m. 1051 (or perhaps a bit earlier), Tostig, d. 1066, earl of Northumbria.
[VE 38 suggests that the marriage took place at the time of Tostig's first exile in 1051, while ASC[D] (s.a. 1051) suggests that Tostig was already married at the time.]

(2) Welf I (IV), d. 1101, duke of Bavaria.
[Gen. Welf. 734; Hist. Welf. 461-2; Ann. Sax. 694]

Children:

by Tostig of Northumbria:
[VE 83]

Children whose names, number, and gender are unknown, apparently still infants in 1065.

See the Commentary section for other alleged children of Judith by Tostig.

by Welf of Bavaria:
[Gen. Welf. 734; Hist. Welf. 461-2; Ann. Sax. 694]

Welf II (V), d. 1120, duke of Bavaria.

Heinrich "der Schwarze" ("the Black"), d. 1126, duke of Bavaria.

Commentary

Supposed sons by Tostig (probably falsely attributed):

Skúli konungsfóstri.

Ketill krókr.

Morkinskinna [276] states that Skúli was a son of Tostig, and mentions Ketill in the same passage without giving any relationship. In a passage describing the same events, Heimskringla [660] states that Skúli and Ketill were brothers of noble English blood, but does not state their parentage. (Some published versions of Heimskringla based on later manuscripts make the brothers sons of Tostig, but this is not in Hollander's translation, based on the earliest manuscripts.) Skúli was evidently an adult at the time of the Battle of Stamford Bridge (1066), and it has also been pointed out [Cleve & Hlawitschka (1982), 27] that the Norwegian names Skúli and Ketill are unknown among Tostig's and Judith's families, so it seems unlikely that they were children of Judith, and it would be extremely difficult to identify them with the children of Judith mentioned above who were still infants in 1065. If these sons have been correctly attributed to Tostig (and that itself is not certain), then they would have been much more likely to have been born by a Norwegian wife or mistress prior to Tostig's marriage to Judith.

—————«»—————

The case for Judith's parentage

The case for making Judith a daughter of count Baldwin IV of Flanders by his second (Norman) wife was first set out by Freeman [3, 663-5], a conclusion that has gained wide acceptance. Because the evidence itself requires some discussion, and because of the circulation among some sources of a more recent (but ultimately unsatisfactory) theory that Judith was a daughter of Richard III of Normandy, the evidence is set out below in some detail, starting with the statements given by the main primary sources.

—————«»—————

Vita Ædwardi Regis
(Harleian MS. 526, a manuscript ca. 1100 of a work written before 1070, probably 1065-1067)
[Barlow (1992), 38; for the date of composition, see also pp. xxix-xxxii, where Barlow argues that Book 1 (from which the quote below was taken) was written 1065-6 (i.e., even before the death of Tostig) and Book 2 in or before 1067]:

"Susceptus est ergo inclitus dux Godwinus ab ipso comite Balduino cum magno honore, tum pro antique federationis iure, tum pro multorum ipsius ducis beneficiorum uicissitudine. Acciderant hec in ipsis nuptiis filii sui ducis Tostini, quando sortitus est uxorem Iuthittam, neptem ipsius clarissimi regis Ædwardi, et sororem predicti comitis Baldewini."

[See also ibid., 82, which mentions Tostig's infant (literally, unweaned) children ("lactentibus liberis"), and repeats the information that Tostig was the husband of Baldwin's sister. In both cases, the count Baldwin referred to was clearly the count then reigning (Baldwin V)]

Table 1. Judith's relations according to Vita Ædwardi:

[The word neptis usually means niece or granddaughter, but can also have the more general meaning of relative.]

—————«»—————

Ordericus Vitalis
Orderic's Ecclesiastical History (written 1123×1141)
[see PL 188, Forester (1853-6); the chapters are numbered differently in Migne and Forester.]

Book 3, Chapter 17 [Migne 188: 284; Chapter 11 in Forester] states that Tostig's father-in-law was Baldwin of Flanders, and that Tostig and William married two sisters: "Exsul igitur Tosticus Flandriam expetiit, ibique socero suo Balduino Flandrensium satrapae Judith uxorem suam commendavit, deinde festinus Normanniam adiit, et Willelmum ducem, cur perjurum suum regnare sineret, fortiter redarguit, seque fideliter, si ipse cum Normannicis viribus in Angliam transfretaret, regni decus obtenturum illi spopondit. Ipsi nempe jamdudum se invicem multum amaverant, duasque sorores, per quas amicitia saepe recalescebat, in coujugio habebant."

Books 4 and 7 list the children of count Baldwin of Flanders, with both lists agreeing (see Table 2):

Book 4, Chapter 14 [Migne 188: 339; Chapter 8 in Forester]: "Balduinus enim gener Rodberti regis Francorum, fortissimus Flandrensium satrapa fuit, et ex Adala conjuge sua filios et filias plures multimoda indole pollentes habuit. Rodbertus Fresio, Arnulfus, Balduinus, Udo Treverorum archiepiscopus et Henricus clericus et Mathildis regina atque Judith uxor Tostici comitis soboles Balduini et Adelae sunt."

Book 7, Chapter 2 [Migne 188: 515; Chapter 1 in Forester]: "Adala vero, filia Rodberti regis, Balduino satrapae Flandrensi in conjugo data fuit, quae multiplicem ei prolem peperit, Rodbertum Fresionem, Arnulfum et Balduinum, consules; Udonem Treverensium archiepiscopum, et Henricum clericum; Mathildem quoque reginam Anglorum, et Juditham Tostici ducis uxorem."

Table 2. Judith's relations according to Orderic Vitalis:

[Note: Orderic makes a number of mistakes here. Arnulf was actually a grandson of the elder Baldwin, and son of the younger Baldwin [see, e.g., GCF, passim]. As pointed out by Freeman [3, 664], archbishop Eudes of Treves was not even a member of this family, being of Alemannian descent, a son of count Eberhard and his wife Ida [Gest. Trev. 183]. Henry the clerk is otherwise unknown.]

—————«»—————

Annalista Saxo
(ca. 1150-60)
[MGH SS 6, 694 (s.a. 1066)]:

"Huius Haroldi coniunx, amita Rodberti comitis de Flandria, ex cognatione beati Ethmundi, regis fuit. Hanc postea Welphus, filius Azzonis marchionis Italorum, duxit uxorem, genuitque ex ea duces Welphum iuniorem et Heinricum. Hec Iudhita dicta fuit, ..."

Table 3. Judith's relations according to Annalista Saxo:

[Note: There was a king Edmund ("Ironside") of EngIand who died in 1016. However, given the term beatus (evidently meaning "blessed" here), it appears from the context that "Ethmundi" is a mistake for "Edwardi" (i.e., Edmund's half-brother Edward the Confessor). As with many German sources, Annalista Saxo here confuses Tostig with his brother king Harold II of England, and erroneously makes Judith a wife of the latter. The same error confusing Tostig and Harold is repeated in Annalista Saxo s.a. 1126 (ibid, 764), with no further information on Judith's parentage. The dotted lines show the relationship if we interpret the word amita by its usual meaning of paternal aunt. Shortly after the first version of this page was uploaded, Chris Phillips pointed out something that had apparently been previously overlooked, i.e., that the identity of "NN" would depend on which count Robert of Flanders Annalista Saxo was mentioning in this reference. If count Robert I "the Frisian" (d. 1093) was intended, then "NN" would be count Baldwin V (in agreement with Vita Ædwardi Regis), whereas if count Robert II (d. 1111) was intended, then "NN" would be Robert the Frisian (in agreement with, Orderic, Alberic, etc.). Since this entry occurs in Annalista Saxo under the year 1066, the first of these alternatives seems more likely.]

—————«»—————

Alberic of Trois Fontaines
Chronica (ca. 1250)
[MGH SS 23, 792]:

"Soror autem eiusdem regis Henrici Adela predicto Balduino Insulano peperit Balduinum huius nominis sextum, Robertum cognomento Fresonum, Philippum patrem Guilelmi de Ypra et filias duas, Iudith, quam duxit Tostinus comes Nortdanimbrorum in Anglia, et Mathildem predictam Normannorum ducissam."

Table 4. Judith's relations according to Alberic of Trois Fontaines:

[Note: Philip was in fact a son, and not brother, of Robert the Frisian. See GCF 324]

—————«»—————

"Florence" of Worcester also makes Judith a daughter of Baldwin V, as he states that Tostig was married to Judith, daughter of a count Baldwin of Flanders who was by context the count Baldwin then reigning [Forester (1854), 152 (s.a. 1051)]. Other sources confirm the Flanders marriage without giving the exact link. So, for example, the "D" manuscript of the Anglo-Saxon Chronicle, discussing the outlawing of Godwin and his sons in 1051, gives only the vague information that Tostig was married to a relative ("mage") of count Baldwin ["... Tostig & his wif, Baldewines mage æt Brycge, ...": ASC[D] 71]. The twelfth century Genealogia Welforum states only that Judith was the daughter of a count of Flanders [Gen. Welf. 734], while the related Historia Welforum Weingartensis states that Judith's father was count Baldwin of Flanders, without identifying which Baldwin [Hist. Welf. 461-2]. The Flemish sources, such as Genealogiae Comitum Flandriae [GCF], do not show any awareness of Judith's existence.

—————«»—————

It is clear that the primary sources cited above cannot all be correct. The most important difference is that some of the sources make Judith a daughter of Baldwin V of Flanders (which would make her a sister of Matilda of Flanders, wife of William the Conqueror), while others make her a sister of Baldwin V (and therefore a daughter of Baldwin IV and an aunt of Matilda). As already mentioned above, the most notable early discussion of the primary evidence was by Freeman [3, 663-5], who concluded that she was a daughter of Baldwin IV by his second Norman wife, and with a few exceptions, there has been wide agreement that Freeman was correct in his analysis.

The most notable alternate suggestion was the one proposed by Decker-Hauff (1956), who argued that Judith was in fact a daughter of Richard III of Normandy and only a step-daughter of count Baldwin of Flanders (and therefore a half-sister of Matilda with the same mother). While this theory was accepted by some German secondary sources [e.g., Wunder (1968) and ES, the latter of which gave the theory wide circulation in computer databases], the serious problems with the Decker-Hauff scenario were pointed out in the detailed arguments of Cleve & Hlawitschka (1982), which supported Freeman's earlier theory. We first discuss the reasons that the Decker-Hauff theory must be rejected, after which the evidence for and against the two obvious Flemish candidates (Baldwin IV and Baldwin V) for the father of Judith will be outlined.

The theory that Richard III of Normandy was Judith's father is weakened from the outset by the fact that the sources giving a parentage to Judith are unanimous in making her the daughter of a count of Flanders, and major evidence would have been needed to overturn the primary sources on this matter. By a major blunder near the beginning of the paper, Decker-Hauff rejected Baldwin IV as Judith's father for a reason that was totally false. Since it would not be feasible to make Judith a daughter of Baldwin IV by his first wife Otgiva of Luxemburg (for otherwise Judith would be a first-cousin once removed of her second husband Welf, see Table 5 below), Decker-Hauff correctly pointed out that the second marriage of Baldwin IV to a daughter of Richard II of Normandy is absolutely crucial to the theory that Judith was a daughter of Baldwin IV. He then rejected that marriage (and with it, the theory that Baldwin IV was Judith's father) by stating that it was no more than a working hypothesis ("Arbeitshypothese") of ninteenth century scholars. Having unequivocally rejected the hypothesis that Baldwin IV was Judith's father, Decker-Hauff then compared the candidacies of Baldwin V and Richard III side-by-side, using the weaknesses in the former to support the latter.

In fact, Decker-Hauff's statement that the second marriage of Baldwin IV was a working hypothesis of the nineteenth century is completely false, for the marriage is attested by the near contemporary testimony of William of Jumièges [see van Houts (1995), 2, 28-9], and Cleve & Hlawitschka [p. 23] pointed out that the marriage is also given in one recension (Continuato Leidensis et Divionensis) of the Flemish source Genealogiae Comitum Flandriae [MGH SS 9, 307], a citation that has apparently been overlooked by other authors (although it should be pointed out here that this source is not necessarily independent of William of Jumièges). Thus, since Decker-Hauff's reason for rejecting the Norman marriage of Baldwin IV was not a valid one, the case for Baldwin IV as Judith's father was not given the hearing that it deserved in Decker-Hauff's remaining arguments. In fact, every reason given by Decker-Hauff for preferring his own theory to the Baldwin V theory could be applied with equal or greater weight for preferring Baldwin IV as the father, and without the enormous disadvantages of the Richard III theory, which must be rejected because of the lack of reasonable evidence.

As for the relative merits of the theories making Baldwin IV or Baldwin V the father of Judith, the following is a brief outline of some of the main points favoring Baldwin IV:

• Vita Ædwardi Regis, the only contemporary source for the parentage of Judith (since ASC does not state her exact relationship to the Flemish counts), states that Judith was a sister of Baldwin V. Even without the other pieces of evidence, this fact alone is a strong point in favor of Freeman's conclusion.

• Two of the early authors who assign Judith as a daughter of Baldwin V, Orderic Vitalis and Alberic of Trois Fontaines, assign children to Baldwin V that do not belong, undermining the reliability of both of these authors with regard to the family of Baldwin V.

• If Judith were a daughter of Baldwin V, then she would have been a granddaughter of Otgiva of Luxemburg, a sister of Welf's maternal grandmother, which would then make Welf and Judith second cousins, related in the third degree (3:3) of consanguinity (see Table 5 below). While not a decisive argument against this affiliation for Judith, it is a definite weakness of the theory making Baldwin V her father. On the other hand, a child of Baldwin IV's Norman marriage would not have this Luxemburg connection, so this consanguinity argument presents no problem for the Baldwin IV theory. [Note added 14 October 2006: This argument depends on the common assumption that Otgiva was a daughter of Frederick of Luxemburg, which cannot be regarded as solidly documented. If, as would seem possible, Otgiva was actually a daughter of Frederick's brother Giselbert, that would weaken the force of this argument.]

• Both Vita Ædwardi Regis and Annalista Saxo indicate that Judith had a close relationship to the Anglo-Saxon royal family. (The term neptis in VE need not be given the literal interpretation of "niece") The theory that Judith was a daughter of Baldwin IV by his Norman wife explains this very well, because Emma of Normandy was the mother of Edward the Confessor (see Table 6). However, a child of Baldwin V would have no obvious close relationship to the English royal family, making the accounts of Vita Ædwardi Regis and Annalista Saxo difficult to explain. (The counts of Flanders were descendants of Ælfred the Great of Wessex, but this descent is too remote to be a plausible explanation of the stated relationship between Judith and the Anglo-Saxon kings.)

• Any children of Baldwin IV by his Norman wife would have been born in the interval 1030×1036, and would therefore have been roughly the same age as the children of Baldwin V (and perhaps even brought up in the same household after the death of Baldwin IV). This makes it easy to explain why writers of the next century or later might have accidently assigned Baldwin IV's child by his Norman marriage as a child of Baldwin V.

• The Norman wife of Baldwin IV was a daughter of Richard II of Normandy by his wife Judith of Brittany, and this provides a good explanation of where Judith received her name (i.e., from her maternal grandmother) if Judith is placed as a daughter of Baldwin IV's second marriage. The theory that Baldwin V was Judith's father does not share this advantage of easily explaining the origin of Judith's name. (Baldwin V was a descendant of the famous Judith of France, wife of Baldwin I, but that was a remote descent.)

Thus, Judith was a daughter of Baldwin IV of Flanders by his wife (name unknown) who was a daughter of Richard II of Normandy. For a more detailed recent discussion of the evidence for Judith's parentage, see Cleve & Hlawitschka (1982). The question of Judith's parentage has also been a topic of discussion in the Usenet newsgroup soc.genealogy.medieval, and more details can be found in the postings there on this subject. Some of the material here is based in part on preliminary discussions that were posted there.

Table 5. The "3:3" consanguinity that would exist between Welf and Judith if she were a daughter of Baldwin V of Flanders as some sources state:

[Note that this consanguinity would be absent for any children of Baldwin IV by his second wife.]

See also the additional note added above regarding this table.

—————«»—————

Table 6. The close relationship between Judith and Edward the Confessor.

[Note that this close relationship would be absent if Judith were placed as a daughter of Baldwin V.]

Bibliography

Ann. Sax. = G. Waitz, ed., Annalista Saxo (MGH SS 6, 542-777)

ASC[D] = G. P. Cubbin, ed., The Anglo-Saxon Chronicle - A Collaborative Edition (Volume 6: MS D) (Cambridge, 1996)

Barlow (1992) = Frank Barlow, ed. & trans., The Life of King Edward (2nd ed., Oxford, 1992).

Cleve & Hlawitschka (1982) = Hartwig Cleve & Eduard Hlawitschka, "Zur Herkunft der Herzogin Judith von Bayern (†1094)", in Fried & Ziegler, eds., Festschrift für Andreas Kraus (Münchener Historische Studien Abteilung Bayerische Geschichte 10, 1982), 15-32.

Decker-Hauff (1956) = Hansmartin Decker-Hauff, "Zur älteren Geschichte der Welfen", in Weingarten, Festschrift zur 900-Jahr-Feier des Klosters 1056-1956, (Weingarten, 1956), 31-48.

Decker-Hauff (1957) = Hansmartin Decker-Hauff, "Judith von der Normandie", in Schwäbische Lebensbilder 6 (1957), 16-27.

ES = Schwennicke, ed. Europäische Stammtafeln.

Forester (1853-6) = Thomas Forester, trans., The Ecclesiastical History of England and Normandy by Ordericus Vitalis (4 vols., London, 1853-6).

Forester (1854) = Thomas Forester, trans., The Chronicle of Florence of Worcester (London, 1854).

Freeman = Edward A. Freeman, The History of the Norman Conquest of England (5 vols. + index vol., Oxford, 1870-9)

GCF = L. C. Bethmann, ed, Genealogiae Comitum Flandriae (MGH SS 9, 302-336; Genealogiae Comitum Flandriae is a collective name for several closely related works giving the genealogy of the early counts of Flanders)

Gen. Welf. = Genealogia Welforum (MGH SS 13, 733-4)

Gest. Trev. = G. Waitz, ed., Gesta Treverorum (MGH SS 8, 111-200)

Heimskringla = Lee M. Hollander, ed. & trans., Heimskringla (Austin, 1964).

Hist. Welf. = Ludwig Weiland, ed., Historia Welforum Weingartensis (MGH SS 21, 454-472)

Luard (1858) = Henry Richards Luard, ed., Lives of Edward the Confessor (Rolls Series 3, London, 1858).

MGH SS = Monumenta Germaniae Historica (Scriptores Series).

Morkinskinna = Theodore M. Andersson and Karl Ellen Gade, eds. & trans., Morkinskinna The earliest chronicle of the Norwegian kings (1030-1157) (Islandica 51, Ithaca, NY, 2000).

Orderic = Orderic Vitalis's Ecclesiastical History [see PL 188, Forester (1853-6)]

PL = Migne, J. P. Patrologiae Cursus Completus (Latin series).

van Houts (1995) = Elisabeth M. C. van Houts, ed. & trans., The Gesta Normannorum Ducum of William of Jumièges, Orderic Vitalis, and Robert of Torigni (2 vols, Oxford, 1995).

VE = Vita Ædwardi Regis (The Life of King Edward [the Confessor]); see Luard (1858) [Latin text only] and Barlow (1992) [Latin text with parallel English translation, the edition quoted here].

Wunder (1968) = Gerd Wunder, "Wilhelm der Eroberer und seine Verwandten in der Sicht der kontinentalen Dynastengenealogie", Genealogisches Jahrbuch 6/7 (1967), 19-41; 8 (1968), 143

Acknowledgements

James Hansen, Chris Phillips, and Henk Verdonk provided information from or copies of sources not available to the compiler.

Chris Phillips pointed out the ambiguity regarding count Robert in Annalista Saxo.

Compiled by Stewart Baldwin
Uploaded 7 October 2002
Revised 11 October 2002 (corrected typos, and added remark about the ambiguity regarding count Robert in Annalista Saxo)

Project Home Page

Henry II Ancestor Table

Alphabetical Index of Ancestor Pages