. 9
( 55 .)


Preuves, cols. 36, 102±3, 431, 433, 440, 466, 474, 487; Actes inedits, nos. xxv and xxix.

Preuves, col. 127.

Ducal Brittany, 1066±1166
to Matthew, his brother, to administer on his behalf with the aid of the
bishop, Benedict, who was their paternal uncle, while he concentrated
his own efforts on restoring ducal authority in Rennes. This is plausible,
but there seems to me no basis for the assertion that Matthew did not
receive any hereditary right in Nantes, and it ignores the signi®cance of
his name. The chronicles and diplomatic sources cited above demon-
strate that Hoel I granted Nantes to his younger son, that Matthew was
accorded the title `comes Nannetensis' (or equivalents), and that he
authorised dispositions of property by landholders of the county of
Nantes, and himself made grants of land within the county, without
requiring the consent of his elder brother.57 Hoel must have intended
Matthew to continue the dynasty of the counts of Nantes, which would
continue to be ruled independently of the rest of the duchy. It was pure
chance that Nantes reverted to the senior line upon Matthew's death
without issue, around 1103.58
After nearly ®fty years of union with the rest of the duchy, the
county of Nantes once again became contentious in the succession crisis
following the death of Duke Conan III in 1148. There ought to have
been no dif®culty about the succession. Conan left a son apparently ®t
to succeed him, but within a year or so after the duke's death, his son
Hoel was acknowledged only as count of Nantes. His sister Bertha and
her husband, Eudo de Porhoet, based at Rennes, were acknowledged as
duchess and duke jure uxoris throughout most of Brittany.59
This extraordinary turn of events requires some explanation. Con-
temporary Breton annals record that, Conan III having disowned Hoel È
as his son (`suum esse ®lium Conanus abnegaverat'), by popular will
Hoel succeeded as count of Nantes.60 This was elaborated by Pierre Le
Baud, to the effect that Conan was persuaded on his deathbed that Hoel È
was not his son and disinherited him. The readiness with which this
version has been accepted and repeated in the historiography is no
doubt due to the fact that it impugns the character of Conan III's wife,
Matilda, an illegitimate daughter of King Henry I, and thus satis®es both
the anti-English and misogynistic sentiments of Le Baud's successors.61
The sheer suf®ciency of this tradition has prevented historians from
examining the succession to Conan III more closely.

Chedeville and Tonnerre, Bretagne feodale, p. 65; Preuves, col. 36.

Breton annals give the date of Matthew's death variously as 1101, 1103 and 1104 (Preuves, cols.

36, 102±103 and 151, cf Preuves, col. 775).
Preuves, cols. 622±4 (Rennes); RT, ii, 6 (eastern Cornouaille).

Preuves, col. 103.

Le Band, Histoire de Bretagne, p. 103. E.g., de la Borderie, Histoire de Bretagne, pp. 42, and

269±72. The strength of the tradition is indicated by its repetition in Chedeville and Tonnerre,

Bretagne feodale, p. 72.

Brittany and the Angevins
In 1908, le vicomte Charles de la Lande de Calan reviewed the
evidence and suggested that Hoel was Conan's illegitimate son.62 This
theory has some appeal. It may be argued, for instance, that Conan
intended to provide for his illegitimate son by giving him the county of
Nantes. This is supported by the choice of the name Hoel, which is
associated both with illegitimate sons of Breton dukes and also with the
counts of Nantes. The name was used by the counts of Nantes from the
tenth century, beginning with an illegitimate son of Alan `Barbetorte'.63
The name was given to Duke Hoel I (whose mother was the grand-
daughter of the ®rst Hoel), instead of a name from the stock used by his
paternal ancestors, the counts of Cornouaille, no doubt to reinforce his
title to the county of Nantes. Additionally, there was a precedent for
severing Nantes from the rest of the duchy for the sake of providing for
a son (albeit a cadet, rather than an illegitimate or disinherited son) in
the case of Conan III's uncle Matthew, the younger son of Hoel I. It is
arguable that Conan III named his son Hoel both because he was
illegitimate (recalling his ancestor, the illegitimate son of Alan `Barbe-
torte') and to add weight to his plan to install him as count of Nantes.
La Lande de Calan's article was a welcome exercise in reviewing the
evidence for the succession crisis, but a more satisfactory explanation is
that advanced by Katharine Keats-Rohan, that Conan III disinherited
his (legitimate) son in the interests of unifying the duchy through the
marriage of his daughter, Bertha, to Alan, earl of Richmond.64 On this
interpretation, Hoel's legitimacy or otherwise is not in issue, and indeed
the original annal-record does not comment on Hoel's parentage,
merely that Conan disowned him, which was tantamount to disinher-
iting him.
The most cogent evidence for this theory is an obituary notice for
Alan, earl of Richmond, which commemorates his attempt to reunite
Brittany. Other evidence is an 1138 charter of Conan III, concerning
property in Nantes, recording the consent of Alan `gener meus', but
making no mention of Hoel.65 Contrary to the traditional death-bed
disinheritance, this arrangement was certainly made some years before
Conan's death, perhaps even before Hoel was born. Extraordinary as it
may seem, in view of the strength of the custom of male succession, a
similar arrangement occurred almost contemporaneously in the county

C. de la Lande de Calan, `Melanges historiques, xix. Le duc Hoel II', Revue de Bretagne 40
62 È
(1908), 180±3.

Chedeville and Tonnerre, Bretagne feodale, pp. 29±31.

K. S. B. Keats-Rohan, `Le role des Bretons dans la politique de colonisation normande de

l'Angleterre (vers 1042±1135)', MSHAB 74 (1996), 181±215 at 205, note 98.
Preuves, cols. 5, and 576±7.

Ducal Brittany, 1066±1166
of Namur. In the 1130s, Godfrey, count of Namur, disinherited his son
Henry the Blind, and gave Namur in marriage with his daughter to
Baldwin IV, count of Hainault, thus uniting the two counties, while
Henry was given a life-interest in Namur.66
Such an ambitious policy required sacri®ces. Hoel was obliged to
sacri®ce his claim to the ducal title in favour of his sister. He is not
known to have married, and his only known child became a nun at
Saint-Sulpice-la-Foret.67 In view of the signi®cance of the name Hoel
à È
outlined above, and the Namur precedent, it may be that Conan
intended to compensate his son with the county of Nantes for his life.
Indeed, the subsequent con¯ict between Hoel and Bertha may have
been limited to a dispute about the degree of Hoel's independence as
count of Nantes.
Similarly, for Alan to succeed to the lordship of Penthievre meant
that one or both of Alan's brothers would have to designate him as their
heir. In the 1120s, Stephen of Penthievre had divided his lands between
his three sons; the eldest, Geoffrey Boterel II, received the eastern
portion (henceforth known as Penthievre or Lamballe), the youngest,

Henry, received the western portion (Treguier or Guingamp), and
Alan, the middle son, received the English lands, the honour of
Richmond. On this basis, Alan had no hereditary right to any of the
Penthievre lands in Brittany. Geoffrey Boterel evidently was not
compliant, as is indicated by his active support for the Empress Matilda
in the English civil war, while Alan fought on the side of King Stephen.
The youngest brother, Henry, on the other hand, seems to have been
persuaded to sacri®ce his independent and potentially hereditary posses-
sion of Treguier in favour of Alan, and to remain unmarried.68 In 1145,

both Alan and Henry were at Conan III's court at Quimper, when Alan
con®rmed their father's grants to a priory in Guingamp, indicating
Alan's lordship of Treguier.69

In fact, Alan predeceased his father-in-law by two years, bringing
Conan's scheme of reuni®cation to nought. Alan's death in 1146 meant
that both Hoel's and Henry's sacri®ces were unnecessary. Henry, at
least, must have decided that the deal was off. Aged nearly ®fty, he

married for the ®rst time and henceforth regarded Treguier as his son's

L. Vanderkindere (ed.), La chronique de Gislebert de Mons, Brussels, 1904, pp. 60±2; J. Falmagne,

Baudouin V, comte de Hainaut 1150±1195, Montreal, 1966, pp. 75, 78; L. Vanderkindere, La

formation territoriale des principautes Belges au moyen age, i, Brussels, 1902, p. 308. I am very grateful
to Laura Napran for this information.
Cart. St-Sulpice, no. lviii.

Preuves, col. 681.

Preuves, col. 595.

Brittany and the Angevins
inheritance.70 Hoel, in contrast, does not seem to have seriously
attempted to claim the duchy. The situation was complicated by the fact
that Alan and Bertha had an infant son, the future Duke Conan IV,
who inherited his father's claims to the duchy of Brittany (including

Treguier) and the honour of Richmond. Bertha promptly remarried, to
Eudo de Porhoet, apparently on the basis that he was well suited to
®ght for her son's cause.
By 1155 the balance of power was clearly in favour of Bertha and
Eudo, and Hoel acknowledged that he held the county of Nantes of his
sister. The peace did not last. For reasons which are not clear, but may
have to do with his capitulation to Bertha, in 1156 Hoel was deposed as
count of Nantes. He was immediately replaced as count by Geoffrey,
the younger brother of Henry II. Several chronicles independently
record that the citizens of Nantes chose Geoffrey to be their count.72
This should not be surprising. In view of the circumstances outlined
above, Nantes was culturally more akin to Anjou than to Armorican
Brittany. At the same time, the county of Nantes was extremely
attractive to the counts of Anjou, for both strategic and ®nancial
reasons. From the point-of-view of the Angevin heartland, the fact that
Nantes controlled the mouth of the Loire made it important that it
should be under the political control of the count of Anjou,73 whether
directly or indirectly. It is not so remarkable, then, that in 1156 a cadet
of the comital family of Anjou became count of Nantes and was
accepted by the populace.
Eudo de Porhoet failed to respond to the events occurring in Nantes
in 1156, no doubt because he was by then engaged in a struggle with his
stepson, the young Conan. Conan had grown up in England, where, as
early as 1153, Henry II recognised him as heir to the honour of
Richmond.74 Conan was anxious to enter into his maternal inheritance
in Brittany, and must have demanded that Eudo deliver the duchy to
him. Presumably Eudo refused, because in the summer of 1156 Conan

See below, p. 54.

In 1155, Hoel made a grant of land in the county of Nantes to the abbey of Buzay, with Bertha's
71 È
consent (`Actes de Buzay', no. 9). Similarly, Fontevraud obtained con®rmations of a grant by
Conan III of an island in the Loire from both Hoel (1153) and Bertha (undated) (Preuves, cols.
617, and 624; BN ms latin 5840, p. 120).
Preuves, col. 103; WN, p. 114; RT, i, p. 298. Hoel does not appear again in Breton sources, but
72 È
attested an act of his nephew, Conan IV, in England, c.1164 (BN ms fr. 22362, f. 7; EYC, iv,
pp. 67±8).
J. Gillingham, Richard the Lionheart, 2nd edn, London, 1989, p. 28.

M. Jones, `The house of Brittany and the honour of Richmond in the late eleventh and twelfth


. 9
( 55 .)