English Historical Review Vol. CXXIII No. 505
© The Author [2008]. Published by Oxford University Press. All rights reserved.
EHR, cxxiii. 505 (Dec. 2008)
Advance Access publication on November 5, 2008
doi:10.1093/ehr/cen276
The De Obitu Willelmi : Propaganda for the
Anglo-Norman Succession, 1087 – 88?*
T he De obitu Willelmi is an intriguing text of just 654 words, purporting
to be an account of the deathbed of William the Conqueror, and
describing how he divided his realms between his eldest son Robert
‘ Curthose ’ , who received Normandy, and the younger William ‘ Rufus ’
who became king of England. It survives only as a rare appendage to
the
Gesta Normannorum Ducum
of William of Jumièges and his
continuators, being found in full at the end of a single manuscript,
although the fi rst 52 words are also found in a closely related copy.
1
These, together with a fragmentary and apparently independent
manuscript which lacks any part of the De obitu Willelmi , are the only
extant copies of the B redaction of the Gesta Normannorum Ducum ,
and the two containing the De obitu Willelmi date to the late eleventh
or early twelfth century. This paper re-examines the evidence for the
date and authorship of the De obitu Willelmi , and proposes a context
for it in the aftermath of the Conqueror’s death.
Fourteen of the forty-seven surviving MSS of the Gesta Normannorum
Ducum
are in England, including the three manuscripts of the B
redaction. All these three are of an early date; the other redactions
continued to be copied into the thirteenth century and beyond. B2,
which alone contains the complete text of De obitu Willelmi , is a
Durham manuscript and is known to have been corrected in the early
twelfth century. Palaeographic analysis, comparing it with other Durham
manuscripts, has indicated a date in the last decade of the eleventh
century or the fi rst quarter of the twelfth.
2
Gullick has subsequently
identifi ed Symeon of Durham’s hand in both the text and its annotations,
confi rming its terminus ante quem as Symeon’s death just before 1130.
3
In addition to De obitu Willelmi and some short anecdotes, this redaction
includes a note of the death in 1092 of Nicholas, bastard son of Duke
* I am grateful to my supervisor Professor Nicholas Brooks and to Dr Elizabeth van Houts for
their encouragement and their helpful comments on an earlier draft of this paper.
1 . Part of the B2 manuscript of the Gesta Normannorum Ducum, London, British Library, MS
Harley 491, fos. 3r – 46v; the B1 manuscript, Oxford, Magdalen College MS 73, fos. 70r – 117v; B3,
Oxford, Bodleian Library MS Rawlinson G. 62, fos. 71v – 73r. The classifi cation of these manuscripts
is taken from E.C.M. van Houts, ed., The Gesta Normannorum Ducum of William of Jumièges,
Orderic Vitalis and Robert of Torign i (2 vols, Oxford, 1992), I: Introduction and Books I – IV, xcv –
cxix. De obitu Willelmi is translated in Gesta Normannorum Ducum VII, ed. van Houts, ii, 185 – 191.
2 . A. Sapir and B.M.J. Speet, De Obitu Willelmi: Kritische beschouwing van een verhalende bron over
Willem de Veroveraar uit de tijd van zijn zonen (Amsterdam, 1976), 3 – 4. I am grateful to Dr A. Sapir
Abulafi a for making this text available to me and to Mr Eric Idema for translating it into English.
3 . M. Gullick, ‘ The Hand of Symeon of Durham: Further Observations on the Durham
Martyrology Scribe ’ , in D. Rollason, ed., Symeon of Durham. Historian of Durham and the North.
Studies in North-Eastern History (Stamford, 1998), 14 – 31.
EHR, cxxiii. 505 (Dec. 2008)
1418
THE DE OBITU WILLELMI
Richard III and therefore the Conqueror’s older cousin; it therefore
post-dates this event.
4
An initial dating range for the B2 redaction and
De obitu Willelmi is thus fi xed as 1092 to c. 1125. If, however, as suggested
(but subsequently rejected) by Sapir and Speet, De obitu Willelmi was
not composed as part of the B redaction, but circulated independently
before being attached to it, then its terminus post quem is fi xed only by
the date of the Conqueror’s death in 1087. van Houts considered that
this was a possibility. She proposed that the other additions to the B
redaction are too different in style from De obitu Willelmi to share a
common author. ‘ The B redactor, anxious to include all the information
about the dukes and the kings that he could fi nd, used De obitu Willelmi
but did not write it himself. ’
5
The text itself was fi rst edited by Marx in 1914, who described it as
the product of ‘ an anonymous monk of St Stephen’s Caen ’ because it
contains a description of the Conqueror’s tomb and the epitaph on it.
6
Since then it has had fl
uctuating fortunes in Anglo-Norman
historiography. David used it uncritically in his 1920 portrait of Robert
Curthose,
7
and in 1964 Douglas wrote, ‘ It is a remarkable description,
and while … it must be received with some discrimination, its simplicity
and circumstantial detail command confi dence, and it is not to be set
aside. ’
8
In 1971 Le Patourel said
9
As a precise and matter-of-fact account it carries conviction; whereas
Orderic’s description was written a generation later, his interpretation bears
the stamp of that generation and the whole has been worked over to make a
literary set-piece. It is better to follow the earlier statement, certainly for any
argument based on detail.
This despite the fact that in 1965 both Barlow and Loyn had noted that
certain passages in De obitu Willelmi describing the Conqueror’s personal
habits had been lifted from Einhard’s Life of Charlemagne, the Vita
Karoli Magni.
10
A major study by Engels then revealed that De obitu Willelmi is a
pastiche of two ninth-century sources, Einhard’s Life of Charlemagne
and the life of Louis the Pious by the anonymous ‘ Astronomer ’ (the Vita
4 . Sapir and Speet, Kritische , 18.
5 . Gesta Normannorum Ducum : ed. van Houts, i, lxiv – lxv.
6 . J. Marx, Guillaume de Jumièges, Gesta Normannorum Ducum. Edition Critique (Rouen and
Paris, 1914), 145 – 8.
7 . C.W. David, Robert Curthose, Duke of Normandy (Cambridge, 1920), 40 – 1.
8 . D.C. Douglas, William the Conqueror: The Norman Impact upon England (London, 1964),
371.
9 . J. Le Patourel, ‘ The Norman Succession, 996 – 1135. ’ ante , lxxxvi (1971), 225 – 50 at 232.
10 . F. Barlow, William I and the Norman Conquest (London, 1965), 43, 177ff.; H.R. Loyn, The
Norman Conquest (London, 1965), 193; O. Holder-Egger, ‘ Einhardi Vita Karoli Magni. ’ Monumenta
Germaniae Historica, Scriptores rerum Germanorum , xxv (1911).
1419
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Hludouuici ).
11
Engels demonstrated that the fi rst part of De obitu
Willelmi , describing the Conqueror’s deathbed and arrangements for
the succession, is lifted from Vita Hludouuici , while the latter part is
a series of disconnected extracts from Vita Karoli Magni , describing
his habits and physical appearance. The parallel texts are reproduced
here in Appendix I . Some historians have seen this as destroying the
credibility of De obitu Willelmi as a useful text: Bates said succinctly, ‘ It
is of little value, ’
12
while Barbara English concluded it was
13
merely a piece created by someone who wanted a short biography of the
Conqueror and copied out the most appropriate sources to which he had
access, altering them in the light of the general (and not specialised)
knowledge available to him … As De obitu Willelmi can so easily be
demonstrated to be merely an echo of older texts, then Orderic Vitalis ’
account assumes a greater importance.
It is certainly true that as an ‘ original ’ source for information about the
last days of the Conqueror, De obitu Willelmi now needs to be seen in a
new light. But nevertheless it was written, and copied into a manuscript
of the Gesta Normannorum Ducum . Moreover it does not contain the
most obvious information nor is it copied in the most obvious way
from the most suitable sources. This raises the question of who might
have written it and why. First, though, it will be helpful to explore
the implications of the detailed work on the text itself done by Engels
and by Sapir and Speet in the light of the new edition of the Gesta
Normannorum Ducum by van Houts.
As Engels demonstrated, De obitu Willelmi is largely composed of
extracts from its two models. Of its 654 words, the fi rst 440 describe the
Conqueror’s last days, and of these 330 are from the Vita Hludouuici ,
with a further 77 being the title and essential changes to names, etc. The
second part describes William in 214 words, of which 181 are taken from
the Vita Karoli Magni . Again, 32 of the ‘ new ’ words are essential changes
to accommodate the different people and places involved. Thus, of the
654 words, only 34 are ‘ voluntary ’ changes. Various hypotheses to explain
this close adherence could be suggested, of which the two extremes on
the spectrum are as follows. Either the topoi themselves had widely
understood symbolic value at the turn of the twelfth century and the
texts were so well known by the intended audience that the symbolism
would be enhanced by a close use of the models, or there was a need for
a written account of the last days of the Conqueror and his choice of
11 . L.J. Engels, ‘ De obitu Willelmi ducis Normannorum regisque Anglorum: Texte, modèles,
valeur et origine ’ , in Anon, Mélanges Christine Mohrmann, Nouveau recueil offert par ses anciens
élèves (Utrecht/Anvers, 1973), 209 – 55.
12 . D. Bates, William the Conqueror (London, 1989), 180.
13 . B. English, ‘ William the Conqueror and the Anglo-Norman Succession ’ , Historical Research ,
lxiv (1991), 221 – 36 at 227.
EHR, cxxiii. 505 (Dec. 2008)
1420
THE DE OBITU WILLELMI
successors, meeting the expected formulae without necessarily adhering
too closely to the facts, and these models supplied a useful short cut.
The generous use of the Charlemagne model fi ts well with both
hypotheses. Charlemagne’s legendary status was widely known in both
lay and ecclesiastical milieux, and Einhard’s text was readily available:
the most recent study reveals 134 surviving manuscripts, of which 24 are
eleventh century or earlier.
14
Any ruler might be compared with
Charlemagne, and especially one such as William the Conqueror, for
whom at least some of the parallels were accurate. Similar comparisons
are made, for example, in the Carmen de Hastingae Proelio : ‘ promptior
est Magno largior et Carolo ’ and by William of Malmesbury, who quotes
the
Vita Karoli Magni
several times.
15
The Conqueror endowed
monasteries generously and reformed the church in Normandy and
England, and had undoubtedly built up a greatly increased ‘ empire ’ .
But despite its superfi cial suitability, there are some curious features
of the use of the Vita Karoli Magni . First, as Engels showed, the selected
passages are used in order, with one exception where a small phrase of
chapter 22 is inserted after the sentence from chapter 25. This is also the
only place where the text is not followed exactly, voce clara being replaced
by voce rauca in De obitu Willelmi . Why should this one change be
inserted out of sequence? Does it neatly serve to lend credence to the
description, by highlighting a well-known feature of the deceased king?
After all, many more people would have heard his voice than would
have been familiar with his domestic habits.
16
Is it signifi cant that the
substituted word occupies the same space on the line, so a copyist would
not be confused by line breaks occurring in different places thereafter?
Or is this to read too much into a simple alteration?
Secondly, the amalgamation of several small extracts from the model
necessitates adjustments where they join. Most of these are managed
easily, but in one place in particular the junction is left unpolished: the
burial place of the Conqueror was already agreed, but in the Vita Karoli
Magni there is a debate about a suitable site for Charlemagne. This is
omitted from De obitu Willelmi , but its echo remains as the copyist
picks up at ‘ At length all were agreed that there was no better place … ’
and inserts ‘ than that which had already been agreed ’ . This slight
14 . M.M. Tischler, ‘ Einharts Vita Karoli; Studien zur Entstehung, überlieferung und Rezeption ’ ,
Monumenta Germaniae Historica, Schriften , xlviii (2001), 20 – 44.
15 . F. Barlow, ed., Carmen de Hastingae Proelio of Guy Bishop of Amiens (Oxford, 1999), 44 – 5,
line 736; William of Malmesbury, e.g.: ‘ Iustae fuit staturae, immensae corpulentiae, fatie fera, fronte
capillis nuda, roboris ingentis in lacertis, ut magno sepe spectaculo fuerit quod nemo eius arcum tenderet
… ’ Gesta Regum Anglorum III. 279; R.A.B. Mynors, R.M. Thomson and M. Winterbottom, eds.,
William of Malmesbury: Gesta Regum Anglorum. The History of the English Kings. Two Volumes
(Oxford, 1998 – 99), i, 508 and see ii, 256 – 8.
16 . There is an echo of this attribute in William of Malmesbury’s Gesta Regum Anglorum III.
281: ed. Mynors et al. , 510 – 11, ‘ it was his practice deliberately to use such oaths, so that the mere
roar from his open mouth might somehow strike terror into the minds of his audience ’ .
1421
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
roughness might suggest that the De obitu Willelmi was produced
carelessly, or in haste.
Thirdly, the model is largely abandoned towards the end, for the
description of the funeral arrangements and epitaph. The role of Rufus
in causing the tomb to be built is stressed, and an epitaph is then given.
But this epitaph differs somewhat from that given by Orderic Vitalis in
his Ecclesiastical History, which Sapir and Speet have established is
almost identical to the version copied down at the tomb in 1522 and
recorded by Charles de Bourgueville in 1588. This could indicate that
Orderic’s is the original, and that the one in De obitu Willelmi is a
corrupted version, but as Sapir and Speet pointed out, it could also
mean that the 1522 copyist relied on Orderic’s Ecclesiastical History to
fi ll lacunae in the engraved inscription, in which case De obitu Willelmi
could contain the original version.
17
In favour of the De obitu Willelmi
version being the original, van Houts has identifi ed another fi fteenth-
or sixteenth-century copy of the epitaph, the same as that in the B2
manuscript,
18
but unless the links between the manuscripts can be
proved, all that can be said is that there was a continuing tradition of
copying the B2-type epitaph in England.
In contrast to the use of the Vita Karoli Magni as a model, the use of
the Vita Hludouuici does not sit comfortably with the fi rst hypothesis,
being neither an obvious nor common topos for the life of a great ruler.
Louis faced three major rebellions by his sons, failed to live up to his
great father’s standards and was once forced to abdicate for a year. If the
source was understood, it was unfl attering; if, as is more likely, it was
relatively unknown in Anglo-Norman society, how readily available was
it for copying, and what prompted the choice?
Twenty-two manuscripts of the Vita Hludouuici survive, but only
fi ve are eleventh century or earlier.
19
There may of course have been
many more in the medieval period, and it is possible that it has survived
less well than the Vita Karoli Magni . But nevertheless it seems relatively
scarce. Of these fi ve manuscripts, four now have the two texts bound
adjacent to each other, so one problem is readily resolved: even if the
Vita Hludouuici
was not as widely available as Einhard’s Life of
Charlemagne, it does seem to have travelled with it. Moreover the
17 . Sapir and Speet, Kritische , 36; Orderic Vitalis ’ Historia Ecclesiastica VIII.1: M. Chibnall, ed.,
The Ecclesiastical History of Orderic Vitalis. Historia Ecclesiastica. Six Volumes (Oxford, 1069 – 1978),
iv, 110 – 13 [ Oderic Vitalis , ed. Chibnall]; C. de Bourgueville, Les recherches et antiquitez de la ville et
université de Caen et lieux circonvoisins des plus remarquables (Caen: 1588) (cited in Sapir and Speet,
Kritische , 36). Orderic’s version of the epitaph is included in this study at the end of Appendix I ,
for comparison. Although Orderic’s autograph copy of Books VII and VIII had probably
disappeared from the St Evroul library by 1522, a late twelfth-century copy was still at St Etienne
Caen. See Oderic Vitalis : ed. Chibnall, iv, xiii – xv.
18 . Gesta Normannorum Ducum : ed. van Houts, ii, 189, footnote 7; the manuscript is London,
British Library, Cotton Titus A. XIX, fo. 114v.
19 . E. von Tremp, ‘ Theganus Gesta Hludowici imp. et Astronomus Vita Hludowici ’ , Monumenta
Germaniae Historica, Scriptores rerum Germanicarum , lxiv (1995), 33 – 4 and 123 – 33.
EHR, cxxiii. 505 (Dec. 2008)
1422
THE DE OBITU WILLELMI
manuscript of the Vita Hludouuici which von Tremp has identifi ed as
being closest to the text used for De obitu Willelmi is one of these four.
This copy, which originated near Chartres, is the only one with the
same wording as De obitu Willelmi for the list of regalia granted to
Rufus.
20
Here then is a partial explanation for the choice of the
‘ Astronomer’s ’ text as a model: it may have been to hand when the
copyist was making his extracts from the
Vita Karoli Magni and
composing De obitu Willelmi .
The use of the Vita Hludouuici in De obitu Willelmi cannot, however,
be dismissed so lightly. If it does not fi t with the fi rst hypothesis, it still
fi ts the second, and the possible signifi cance of this merits further
investigation. Unlike the Vita Karoli Magni , it has been used (with one
small but very important exception) as a sequence of extracts, selected
from three consecutive chapters, but within this material a quarter has
been changed. The one place where the text has been rearranged has a
crucial effect. The phrase ‘ which with God and the leading men of the
palace as witnesses he had already granted to him a long time previously ’
is carried forward to make it refer directly to the prodigal son (Robert
Curthose), not to the favoured heir (Rufus).
21
The resulting emphasis in this fi rst part of De obitu Willelmi is
distinctive. As Engels noted, there is no mention of the youngest son,
Henry, despite there being a place for him in the ‘ Astronomer’s ’ model.
Word for word, 11% is the title, introduction and conclusion, linking
the description to the Conqueror; 19% concerns his sickness, last rites
and death; 11% his division of the treasury for pious bequests; 10% is a
list of witnesses, only 3% describes the promise of the regalia to Rufus,
while a full 46% describes the rift with Curthose, his unsuitability for
rule and his father’s reluctant agreement to confi rm him as duke of
Normandy. This part alone takes up almost a third of the whole De
obitu Willelmi , suggesting that the major aim of the author was to stress
the Conqueror’s disillusionment with his eldest son. This effect is also
emphasised by placing these extracts at the beginning of De obitu
Willelmi , instead of after the description of the Conqueror in life,
which would be a more conventional order to adopt. Indeed, De obitu
Willelmi does not seem to fi t into any contemporary literary category,
which might point to a particular and unusual motive for its
composition.
20 . E. von Tremp, ‘ Die Uberlieferung der Vita Hludowici imperatoris des Astronomus ’ ,
Monumenta Germaniae Historica, Studien und texte , i (1991), 17 – 19 and 58 – 60; I would like to
thank Mr Charles West for his help translating this text. Of these fi ve early manuscripts, one [Paris
B.N., lat. 5943 A] is the autograph of Ademars of Chabannes, the other four originate, respectively,
in northern France, Trier, Chartres and St-Germain-des-Prés. The manuscript von Tremp identifi ed
as the exemplar for De obitu Willelmi is the early to mid eleventh-century Chartres ‘ P1 ’ manuscript,
Paris B.N. lat. 5354, which includes Vita Karoli Magni on fos. 50r – 61v and VH on fos. 61v – 85v. But
see Gesta Normannorum Ducum : ed. van Houts, i, lxiii – lxiv.
21 . See Appendix I . ‘ quam/quem Deo teste et proceribus palatii ille secum et ante se largitus ei
fuerat ’ .
1423
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
The De obitu Willelmi states clearly that on his deathbed the Conqueror
was at fi rst minded to disinherit Curthose completely, and was reluctantly
persuaded to grant him Normandy, predicting as he did so that it would
soon descend into chaos. Engels suggested that De obitu Willelmi was
composed in England in the fi rst few years of the twelfth century, before
Curthose was deposed at the Battle of Tinchebrai (1106) and when Normandy
was indeed sliding into chaos. Against this, Sapir and Speet decided that the
omission of Henry, despite the mention of a third brother in the Vita
Hludouuici model, must mean that it was composed before 1100.
There is a third possible piece of dating evidence which should be
taken into account here, namely the epitaph with which De obitu
Willelmi ends. Orderic Vitalis names Otto the Goldsmith as the man
chosen by Rufus to make the tomb, and describes how from among
many epitaphs written, that composed by Thomas of York (1070 – 1100)
was selected ‘ because of his metropolitan dignity ’.
22
The combination
of this fl attering emphasis on the archbishop’s poem with the deliberate
omission of Henry, which suggests that he was of no relevance, together
point strongly to a date before 1100.
Sapir and Speet also concluded that since the B redaction of Gesta
Normannorum Ducum refers to the death of Abbot Nicholas in 1092,
this must give the terminus post quem. Their fi rst hypothesis was that De
obitu Willelmi could have been written in the Rouen area at the time of
Rufus’s visit to Normandy between late 1096 and early 1097 (by which
time Curthose was preparing to depart on Crusade or had already left)
and they suggested furthermore that Rufus might then have built the
tomb for which he is commended in De obitu Willelmi . Their second
hypothesis, which they subsequently rejected, was that De obitu Willelmi
might initially have circulated independently from the
Gesta
Normannorum Ducum (in which case the 1092 terminus post quem
disappears) and they suggested rather unconvincingly that De obitu
Willelmi might in such a case have been produced in association with
the Treaty of Rouen in 1091, at which Curthose and Rufus mended their
quarrel and then took arms against Henry.
23
If, however, De obitu Willelmi was written separately and only added
to the end of the B2 manuscript later, the terminus post quem is actually
September 1087, the date of the Conqueror’s death. Several factors
should be considered here. First, there is the curious fact that the Vita
Hludouuici manuscript closest to that used for De obitu Willelmi is from
near Chartres, but all three surviving manuscripts of the B redaction of
Gesta Normannorum Ducum are in England, as are the two fourteenth-
century copies of excerpts from it.
24
There is no continental tradition of
22 . Oderic Vitalis VIII.1: ed. Chibnall, iv, 110 – 11.
23 . Engels, De obitu Willelmi ducis , 253 – 5; Sapir and Speet ( Kritische ), 30 – 4.
24 . Cambridge, Corpus Christi College, MS 138, 167 – 77, and London, College of Arms, MS1,
fos. 176r – 179v. Gesta Normannorum Ducum : ed. van Houts, i, cxix.
EHR, cxxiii. 505 (Dec. 2008)
1424
THE DE OBITU WILLELMI
including
De obitu Willelmi
in the
Gesta Normannorum Ducum .
Secondly, even in England there is no surviving evidence for a continuing
tradition: there are later copies of the Gesta Normannorum Ducum of
English provenance, but none include De obitu Willelmi .
25
Thirdly, De
obitu Willelmi does not generally seem to have been employed as a
source for later accounts of the Conqueror’s death, even in England.
There may be faint echoes of its assertion that he had to be persuaded
to divide his lands, in the Gesta Regum Anglorum , but the latest edition
emphasises the similarity somewhat unwarrantedly, rendering
‘ Normanniam inuitus et coactus Rotberto, Angliam Willelmo … delegauit ’
as ‘ Reluctantly and under pressure he entrusted Normandy to Robert;
England he gave to William ’ .
26
William of Malmesbury was writing in
the reign of Henry I, while Curthose was in prison, so the stress achieved
by the word order here is not too surprising, but the addition of the
extra English verb lends it additional weight. A second possible slight
echo of De obitu Willelmi ’ s detail is in John of Worcester (and following
him Symeon of Durham), which at this point in the narrative includes
material not in the Anglo Saxon Chronicle, notably the release of Odo,
and also omits Henry. Otherwise, De obitu Willelmi ’ s details are not
shared with other sources.
Overall, the impression is that De obitu Willelmi , and perhaps also
the B2 redaction of the Gesta Normannorum Ducum , is something of a
dead end, divorced from the main stream of the Gesta Normannorum
Ducum and the wider tradition of historical writing. It is a text which
now seems geographically removed from its supposed Norman origins,
which has not been copied in the way that other manuscripts in the
Gesta Normannorum Ducum family have, and which does not appear to
have infl uenced subsequent authors. Wace, writing in Normandy in the
third quarter of the twelfth century, uses some of the anecdotes which
are interpolated into the B redaction of Gesta Normannorum Ducum ,
but includes nothing which suggests that he was familiar with the De
obitu Willelmi
. Indeed the whole tenor of his description of the
Conqueror’s deathbed is markedly different from that in De obitu
Willelmi , and a more likely source here is Orderic Vitalis.
27
It is unlikely
that in the middle ages, De obitu Willelmi would have been set aside as
an unreliable blending of its two ninth-century sources, since it is merely
25 . For example, a fourteenth-century copy of the D redaction from Whalley Abbey, Lancashire
(Cambridge, Trinity College MS O.1.17, fos. 212v – 244r) and a fi re damaged and now divided late
twelfth-century copy of the F redaction from Reading Abbey, which was listed in the 1191 × 93
Cartulary of Reading (London, British Library, MS Cotton Vitellius A. VIII fos. 5r – 100v +
Cambridge, Gonville and Caius, MS 177/210), van Houts nos D4 and F7; Gesta Normannorum
Ducum : ed. van Houts, i, cii – ciii and cxii – cxiii.
26 . Gesta Regum Anglorum III.282: eds. Mynors et al. , i, 510 – 11.
27 . E. van Houts, ‘ Wace as Historian ’ , in G.S. Burgess and E. van Houts, eds., The History of
the Norman People: Wace’s Roman de Rou (Woodbridge, 2004), xxx – lxii, xxxviii; Oderic Vitalis , ed.
Chibnall, iv, xxi – xxii.
1425
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
an extreme example of the common practice of drawing from older
models and topoi. Contemporary writers might rather have been
impressed by the parallels made. An alternative explanation is that it
had served its particular purpose by the early twelfth century, and it
was only by chance that it survived at all.
There is nothing here which means that De obitu Willelmi must have
been included in the B redaction of Gesta Normannorum Ducum from
the beginning, and some justifi cation for suggesting that it might have
been added later. If this were the case, De obitu Willelmi could be an
independent text, composed at some point in the reign of William
Rufus, aiming primarily to discredit Curthose but also assuming that
Henry was of no account. Since Henry was more important once
Curthose had left for the Crusade,
28
this supports the idea that De obitu
Willelmi was composed before 1097. This would mean a return to the
old view of it as much the earliest account of the death of the Conqueror,
signifi cantly earlier than the date suggested by Engels, and probably
earlier than Sapir and Speet’s favoured date of 1096 – 97. van Houts
proposed that De obitu Willelmi predates 1100, and called it ‘ a piece of
propaganda in favour of King William Rufus, too different from the
anecdotes [inserted in the B redaction] to come from the same pen ’ . If
this is so, there is no reason why it could not have circulated independently
for some time before it was incorporated.
29
If one accepts that De obitu Willelmi was written within 13 years of
the Conqueror’s death, and probably within a decade, it achieves a new
importance, not necessarily for the surface facts it presents, but for the
light it might shed on the situation in the fi rst part of Rufus’s reign. The
remainder of this study will consider why and by whom De obitu
Willelmi might have been written, in order to see if a coherent hypothesis
for its existence can be achieved.
As Engels stressed, a key to understanding De obitu Willelmi is to
consider the places where it differs from its models, since these are likely
to result from deliberate choice.
30
An obvious feature is the men named
in the text. A comparison of those named at the deathbed in the Vita
Hludouuici with those in De obitu Willelmi shows that only three
‘ unnecessary ’ changes were made. The younger son of the king was
omitted, and two court offi cials, John ‘ medicus ’ and Gerard ‘ cancellarius ’
were inserted. These changes are all the more striking in view of the
28 . For example in 1096, ‘ Count Henry went over to King William, whose loyal adherent he
became. The king then granted him the whole of the counties of the Cotentin and Bayeux, except
for the city of Bayeux and the town of Caen. ’ Gesta Normannorum Ducum VIII.7: ed. van Houts,
ii, 210 – 13. Henry did not, however, witness many of Rufus’s acta . H.W.C. Davis, ed., Regesta
Regum Anglo-Normannorum 1066 – 1154. Three Volumes. Volume i. Regesta Willelmi Conquestoris et
Willelmi Rufi , 1066 – 1100 (Oxford, 1913).
29 . Engels, De obitu Willelmi ducis , 253 – 4; Sapir and Speet, Kritische , 31; Gesta Normannorum
Ducum : ed. van Houts, i, lxiv – lxv.
30 . Engels, De obitu Willelmi ducis , especially at 234.
EHR, cxxiii. 505 (Dec. 2008)
1426
THE DE OBITU WILLELMI
otherwise close parallelism of the texts. Orderic by contrast, writing in
Henry I’s reign, included Henry, and omitted both these offi cials. Both
Engels and Sapir and Speet have noted these differences; Sapir and
Speet also gave a summary of the careers of two people who may
correspond to John and Gerard.
The inclusion of these two men, John and Gerard, in such a restricted
description as De obitu Willelmi , is certainly curious and requires some
explanation. Turning to John fi rst, it is not surprising that a doctor is
mentioned at the deathbed, since there were certainly medical
practitioners in Normandy and England at this time. What is signifi cant
is that a John is named, since as will be demonstrated below he is not
identifi able in any other sources relating to the Conqueror.
Table 1 summarises the key information about the nine doctors
associated with the Conqueror in non-narrative sources. Pre-eminent
among the royal doctors in Normandy was Gilbert Maminot, bishop of
Lisieux. A royal chaplain, he was consecrated bishop in 1077. On at
least one occasion, he was with the king in England, witnessing a charter
at Windsor in 1070.
31
Despite the dispute between his monastery and
the bishop, Orderic several times praised Gilbert’s medical abilities,
describing him as ‘ the king’s physician and chaplain … A man of great
learning and eloquence … ’
32
Gilbert also held land in England at
Domesday, in sixteen counties and from several previous holders,
perhaps indicating multiple small gifts from the king. Some of these
holdings may have related to his years as a royal chaplain, for example
his three virgates of the royal demesne at Windsor, where Albert the
Clerk and Eudo the steward also held.
33
While much smaller than the
English estates of Odo of Bayeux or Geoffrey of Coutances, Gilbert’s
total estate (valued at nearly £130) was still signifi cant.
34
Gilbert’s
presence at the deathbed is thus readily explained, especially since the
king was near Rouen, only about 80 km from Lisieux. De obitu Willelmi ,
however, does not mention that he was a doctor, reserving that title for
John. Orderic, while not actually calling Gilbert a ‘ medicus ’ in this
context (perhaps in deference to his episcopal rank), named him as
one of several senior clergy who watched over the king’s ‘ spiritual and
corporal needs ’ .
35
In this company, John’s inclusion in De obitu Willelmi as the only
named doctor is noteworthy. De obitu Willelmi is the only source to
31 . B. 81: D. Bates, ed., Regesta Regum Anglo-Normannorum. The Acta of William I (1066 – 1087)
(Oxford, 1998), 343 – 5.
32 . Oderic Vitalis V.3: ed. Chibnall, iii, 18 – 23.
33 . D[omesday] B[ook] Berks. i.56d.
34 . Odo’s personal holding exceeded £3,000, Geoffrey of Coutances held land valued at just
over £750, while the next largest estate of a Norman bishop was that of Gilbert, Bishop of Evreux,
valued at only £22. DB passim.
35 . Oderic Vitalis VII.14: ed. Chibnall, iv, 80 – 1. Chibnall translates ‘ archiater ’ as ‘ physician ’
here.
1427
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Table
1:
The doctors associated with William the Conqueror. Diplomatic and Domesday evidence
Name
No. of charters TRE
(with dates)
No. of William the Conqueror
acta
(with dating range)
Land held TRE
Land held
1086
a
Preferment?
Baldwin
b
1 (
1062
)
13
(
1066
–
1086
)£
11
£
16
.3
s.
Abbot of Bury St Edmunds,
1064
Gilbert
—
27
(
1042
×
1084
)
—
£
128
.5
s.
Bishop of Lisieux,
1077
Nigel
c
—
3 (
1035
×
1066
)
—
£
41
.6
s.
—
Gontard
d
—
5 (
1066
×
1078
–
c.
1087
) —
4 churches
Abbot of Jumièges,
1078
Robert
e
—
1 (
1081
×
1086
)
—
—
—
Rodolfus
f
—
1 (
1063
×
1066
)
—
—
—
Aelfric
—
—
—
4
acres
—
Tetbald
—
—
—
c.
£
1
—
John
—
—
—
—
Bishop
of
Wells,
1088
a
Some parcels of land are not valued in Domesday, so the fi
gures in the table are approximations only.
b
Baldwin witnessed a grant by Edward the Confessor in favour of Waltham Abbey in
1062
: Keynes (
Regenbald
). His personal landholdings are complex because he held land of
King Edward before becoming abbot, and this had been transferred to St Denis: B.
254
Bates (
Regesta
),
pp.
767
–
9. Moreover it is not always clear if land granted by William the
Conqueror was to Baldwin in person or to him in his capacity as Abbot.
c
The connection between the Domesday
‘ Nigel medicus
’ and the Nigel who witnessed F.
95
, F.
166
and F.
227
is not certain. Fauroux (
Recueil
),
pp.
247
–
8,
357
–
8,
435
–
7.
d
These
churches,
one
of
which
was
valued
at
28
s.
at Domesday, were transferred to St Wandrille when Gontard entered the abbey as a monk between
1066
and
1078
. B.
263
: Bates
( Regesta
),
p.
792
.
e
Named
as
a
subtenant
in
B.
49
: Bates (
Regesta
),
p.
235
.
f
Witnessed
F.
165
: Fauroux (
Recueil
),
p.
357
.
EHR, cxxiii. 505 (Dec. 2008)
1428
THE DE OBITU WILLELMI
mention a John ‘ medicus ’ before the death of the Conqueror, and as
Table 1 demonstrates, no doctor of this name held land at Domesday.
While this does not rule out the possibility that he acted as a royal
doctor, it suggests that he had not yet risen high in the king’s favour. As
already observed, there is no place for a named physician in the Vita
Hludouuici , so the inference must be that this is a calculated decision to
draw attention to him. Leaving aside the possibility that he was in some
way being blamed for the king’s death, or at least being held responsible
(neither of which are likely in this documentary context), this suggests
that he is mentioned in order to involve him in the events in some other
way, or to heighten his reputation.
If he was not offi cially a court ‘ medicus ’ , is there evidence for another
John associated with the Conqueror who might fi t this role? John was
not an unusual name at this period,
36
but there is no surviving evidence
for a chaplain named John among all those who witnessed charters, or
were named in them, during the lifetime of the Conqueror ( Table 2 ).
This material represents a minimum of twenty-two chaplains, but since
few of the individuals have identifying second names, the actual total
could be much greater, up to fi fty-six. In the two cases that do survive
of a ‘ John ’ witnessing a charter, the men concerned were both attached
to the recipient houses.
37
There is additionally one anomalous attestation
‘ Johannis Bathonensis episcopi ’ in a Durham charter in a hand from the
later twelfth century, purporting to date to about 1086; but since this
John was not consecrated until 1088 and only moved the see from Wells
to Bath c. 1091, this must represent a later accretion.
38
The possibility
does remain, however, that in its original form (if it had one) the charter
was witnessed by this John as a chaplain, and his title was later changed
anachronistically.
39
Thirteen names of chaplains occur as Domesday landholders, often
of modest amounts, but John is not among these either. For example,
36 . For example, Bishop John of Avranches (1060 – 7) went on to become Archbishop of Rouen
(1067 – 79), and Fécamp was ruled by an Abbot John from 1028 to 1079.
37 . ‘ Johannes monachus noster ’ in a dispute between Marmoutiers and St Pierre de la Couture:
F. 159 (1063 × 1066), in M. Fauroux, ed., Recueil des actes des ducs de Normandie de 911 a 1066 (Caen,
1961), 344 – 8; ‘ Ex parte sancte Trinitatis … Johannes ’ for La Trinité Fécamp: B. 147 (1066 × 1087),
Bates ( Regesta ), 489. The latter appears to have been identifi ed by Mooers as an attestation by the
John who became bishop of Bath and Wells, and furthermore to have been used as evidence that
this John was a royal chaplain from 1066 to 1087, but there seems to be no justifi cation for these
assumptions. S. Mooers Christelow, ‘ Chancellors and Curial Bishops: Ecclesiastical Promotions
and Power in Anglo-Norman England ’ , Anglo-Norman Studies , xxii (2000), 49 – 69, especially at 57.
38 . B. 115: Bates ( Regesta ), 407 – 8. John made his profession of obedience to Canterbury in July
1088 and was consecrated in the same month. F.M.R. Ramsey, ed., English Episcopal Acta x: Bath
and Wells, 1061 – 1205 (Oxford, 1995). The grant of Bath Abbey to John, enabling him to move the
see from Wells, was confi rmed in January 1091: W. Hunt, ed., ‘ Two Chartularies of the Priory of
St Peter at Bath’, Somerset Record Society , vii (1893), especially at 40 – 2.
39 . Compare, for example, B. 232 (1069), where Herfast is described in a late eleventh-century
cartulary copy of a charter as ‘ Erfast tunc capellani, postea episcopi ’ . Bates, Regesta , 725.
1429
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Table 2: Acta of William the Conqueror in which chaplains are named
a
Name
b
No.
of acta
Date range (with acta at
dating extremes)
Rainald
12
1050 × 1066 (F. 197) – 1080 ×
1084 (B. 162)
Samson
9
1072 × 1085 (B. 265) – 1083
(B. 64)
Odo (Queen’s chaplain)
5
1078 × 1083 (B. 160) – 1083
(B. 64)
Robert
c
5
1035 × 1065 (F. 164) – 1086
(B. 115)
d
Ingelric
4
1066 × 1067 (B. 216) – 1069
(B. 138)
Bernard
2
1068 (B. 181) – 1081 (B. 39)
Henry (Queen’s chaplain)
2
1079 × 1087 (B. 164) – 1080 ×
1083 (B. 161)
Theobald
2
1052 × 1058 (F. 141) – 1055
(F. 137)
Thomas
2
1068 (B. 181) – 1080 × 1084
(B. 162)
Baldwin
1
1052 × 1058 (F. 141)
Gerard
1
1073 × 1077 (B. 173a)
e
Goisfrid
1
1083 (B. 64)
Gontard
f
1
1066 × 1078 (B. 263)
Herfast
1
1069 (B. 232)
Maurice
g
1
1086 (B. 115)
Michael
1
1068 (B. 181)
Osmund
h
1
1074 (B. 27)
Ralf
i
1
1040 × 1050 (F. 117)
Ranulf
1
1086 (B. 115)
Seufredus
1
1042 × 1066 (F. 187)
Walter
1
1086 (B. 181)
William
1
1068 (B. 181)
a
Information drawn from Bates ( Regesta ) and Fauroux ( Recueil ).
b
Multiple entries may refer to more than one individual, of the same name. This table therefore
refers to a minimum of 22 chaplains, and a theoretical maximum of 56.
c
There were at least three royal chaplains called Robert: Robert Losinga became bishop of
Hereford in 1079, Robert de Limesey became bishop of Chester in 1086, Robert Bloet was
appointed to Lincoln by Rufus in 1094, after serving as chancellor.
d
Bates is undecided if B. 115 is genuine. See discussion in Bates ( Regesta ), p. 407.
e
Witnessed as chaplain with Robert Curthose; the same charter was witnessed by Osmund as
chancellor with the Conqueror.
f
The same Gontard who became abbot of Jumièges in 1078.
g
Probably the same Maurice who appears in eight acta as chancellor (1078 – 82).
h
Probably the same Osmund who appears in fi ve acta as chancellor (1067 – 78).
i
There are a further fi ve occurrences of ‘ Ralf ’ with no title ‘ chaplain ’ or ‘ priest’, between
1035 × 1066 (F. 166) and 1070 × 1083 (B. 206).
EHR, cxxiii. 505 (Dec. 2008)
1430
THE DE OBITU WILLELMI
Samson, who is probably the man who became bishop of Worcester in
1096, held land valued at £11 3 s. and appears in nine royal charters, one of
which concerns the donation of land to him.
40
This practice of giving
small grants of land to royal chaplains was a well-established tradition:
several of Edward the Confessor’s chaplains held land in 1066, and some
of them were still landholders in 1086.
41
Chaplains such as Seufredus, who
witness early ducal charters, are not recorded in Domesday and perhaps
had already died, but many of the chaplains appointed to bishoprics by
the Conqueror had personal holdings, perhaps as a reward for their earlier
services.
42
(Bishop Peter of Chester, who died before 1086, even has two
references preserved to previous small landholdings of his.
43
)
There are also numerous lay royal servants who witnessed charters or
appear in Domesday. (Wiltshire Domesday alone lists twenty-nine such
men, including a cook, a doctor and three chamberlains, with an average
holding of £3.4 s. ) Three of these servants were called John, but none
seem likely to have been an important royal doctor: one (with no job
listed) held land to the value of 40s.; John ‘ hostillarius ’ held land in two
counties to the value of £6.5 s. ; and John ‘ camerarius ’ held land valued at
35s. which the queen had given him.
44
By contrast, once Rufus came to the throne, the evidence proliferates
for a royal doctor named John. A priest of the city of Tours, he was
appointed bishop of Wells in summer 1088, with a speed uncharacteristic
of Rufus’s usual treatment of episcopal vacancies. In near-contemporary
sources, this John is always referred to as Turonicus , but, perhaps as a result
of a misreading of Wellensis , he is referred to in Anglia Sacra and thereafter
as John de Villula.
45
Ranulf Flambard is often cited as an example of a
man who rose swiftly by lay employment at Rufus’s court, to become a
bishop. But he already held land at Domesday, valued at over £20, and he
was not consecrated until 1099, after a decade of curial service. Compared
with him, John’s rise from obscurity to the episcopacy was meteoric.
40 . B. 265: Bates, Regesta , 796.
41 . For example, Edward the Confessor’s two chaplains Ingelric, who held land valued at over
£300 at one stage and at Domesday still held property valued at £24, and Regenbald, whose land
at Domesday was valued at about £40, spread over fi ve counties. S. Keynes, ‘ Regenbald the
Chancellor (sic), ’ Anglo-Norman Studies , x (1988), 185 – 222.
42 . For example, Bishop Walkelin of Winchester held a hide of land valued at £4 at Brownwich,
and ‘ it is not of the bishopric ’ : DB Hants. i.40d; Bishop Osbern of Exeter held lands attached to
Bosham church, to a value of £60.15 s. in 1086, and had held them from Edward the Confessor: DB
Sussex i.17b.
43 . Two churches and about two hides of land in Somerset, valued at £3, DB Somerset i.91c;
and a close in the borough of Wallingford rendering 4d., DB Berks. i.56b.
44 . Wimbourne, DB Dorset i.85a; DB Wilts. i.74c and DB Somerset i.87c, i.90c; DB Glos.
i.163d. Compare this last holding with that of ‘ William camerarius ‘ : nearly £60 in eight counties,
including a vineyard in Holborn.
45 . Wharton’s Anglia Sacra I: ‘ The Canon of Wells, ’ 559. D. Greenway, John le Neve; Fasti
Ecclesiae Anglicanae 1066 – 1300. VII. Bath and Wells (London, 2001), 1. See, for example B. 68, an
original charter dated 1072 (Bates, Regesta , 311 – 14) which Giso, John’s predecessor at Wells, attested
as Giso UUellensis.
1431
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Orderic implied, but did not state categorically, that this ‘ John the
doctor ’ acted as a chaplain to Rufus; he lumped him together with those
‘ chaplains and favourites ’ who obtained bishoprics from him, to whom
he ‘ bestowed ecclesiastical honours, like hireling’s wages, on clerks and
monks of the court, looking less for piety in these men than for
obsequiousness and willing service in secular affairs’.
46
Smith, in his
study of John’s career, suggested that Rufus was already using him as a
chaplain before he became king, although there is no evidence that
Rufus had an independent household before 1087.
47
If John’s service
only began then, there was not much time for him to make such an
impression on the young king as would justify his preferment the
following year.
The other chroniclers were clear that the John who became bishop of
Bath and Wells was a doctor, and they were not always complimentary
about his abilities. ‘ John, who originated from Tours, succeeded Giso as
bishop of Wells. He practised as a doctor, though he had not been
trained as one. ’
48
‘ A native of Tours and by practice rather than by book-
learning a skilled physician. ’
49
Thus there is an anomalous situation for the fi rst of the two royal
servants named in De obitu Willelmi . John ‘ medicus ’ not only had no
place in the model used but also had no discernable place in the royal
household prior to 1087. Once Rufus became king, however, a man
who could be this John emerged swiftly into an important role, more
suited to that accorded him in De obitu Willelmi . The possibility of a
causal link between these facts will be discussed later.
Turning now to the second royal servant in De obitu Willelmi , a
chancellor is as plausible an attendant near a royal deathbed as a doctor.
Keynes has shown that from at least Edward, the Confessor’s reign there
was an embryonic chancery in England. A succession of Normans acted
as chancellor, probably beginning with Herfast in about 1068. The
fl uidity of the role and title may be indicated by the presence of a witness
‘ Herfastus capellanus ’ on a charter in 1069.
50
Herfast seems to have been
replaced, in turn, by Osmund and Maurice, as each was elevated to the
episcopate. In each case, surviving royal charters are also witnessed by
chaplains of these names during their ‘ term of offi ce ’ as chancellor.
51
Maurice was consecrated bishop of London in 1086. Davis suggested
that he was succeeded as chancellor by Robert Bloet, but in 1931
46 . Oderic Vitalis , X.2: ed. Chibnall, v, 204 – 5 and 202 – 3.
47 . R.A.L. Smith, Collected Papers (London, 1947), 75.
48 . John of Worcester, entry for 1091: P. McGurk, ed., The Chronicle of John of Worcester. Three
Volumes. Volume iii: The Annals from 1067 to 1140 … (Oxford, 1998), iii, 57.
49 . Gesta Regum Anglorum IV.340: eds. Mynors et al. , 588 – 9.
50 . B. 181, see also B. 138 and 81; for Herfast as capellanus in 1069 (probably April) see B. 232.
Bates, Regesta , 594 – 601, 463 – 5, 343 – 5, 725.
51 . Osmund: see B. 27; Maurice: see B. 110 (a Durham forgery) and B. 115 (possibly genuine).
Bates, Regesta , 176 – 8, 394 – 7, 406 – 8.
EHR, cxxiii. 505 (Dec. 2008)
1432
THE DE OBITU WILLELMI
Galbraith proposed that Gerard, or Girard, afterwards bishop of
Hereford, was briefl y chancellor in between.
52
This was based partly on
the evidence of Hugh the Chantor of York, and partly on two writs,
both attested by ‘ G.cancellario ’ , the fi rst of which Galbraith dated to
late in the Conqueror’s reign or soon after, the second to early in the
reign of Rufus. Bates has since concluded that neither can be dated
more closely than 1086×1088,
53
but it is nevertheless probable that
Gerard served, albeit briefl y, in the chancery at the end of the Conqueror’s
reign. A chaplain named Gerard appears in the witness list of one of the
acta of the Conqueror’s reign (see Table 2 ), and Eadmer described
Gerard not as Rufus’s chancellor but as a
‘
chaplain of the king,
’
complying with his wishes in his dispute with Anselm.
54
A ‘ Gerald the
chaplain ’ held three pieces of land valued together at Domesday at
£3 10 s.
55
As with John ‘ medicus ’ , the Vita Hludouuici model does not justify
the inclusion of Gerard ‘ cancellarius ’ in De obitu Willelmi . Unlike John,
however, there is evidence that a man of this name witnessed royal
charters prior to 1088 and held land at Domesday, and he may have
been chancellor briefl y. While there is no certainty that all this evidence
relates to one and the same person, there is at least a body of evidence
that such a man existed. Unlike John, Gerard continued to act as a royal
chaplain for Rufus, only obtaining a bishopric in 1096. Neither of these
men seems to have held a prominent enough position at court in 1087
to warrant their mention in De obitu Willelmi unless there were some
particular motive for doing so. Of the two, John’s inclusion is by far the
more curious. The question therefore arises whether either of them was
linked in some specifi c way to Rufus around the time of his accession,
or even to the production of De obitu Willelmi , as van Houts has
proposed.
Superfi cially, De obitu Willelmi is not greatly concerned with the
English succession. The part of the Vita Hludouuici model dealing with
the royal succession is shortened, leaving only the non-committal
statement ‘ he allowed his son William to have the crown, the sword and
the golden sceptre with inlaid jewels’. In contrast, extra phrases are
added to emphasise the Conqueror’s fears for Normandy if Curthose
should become duke. When these facts are combined with the probable
dating of
De obitu Willelmi
to Rufus’s reign, and the surviving
manuscripts being in England, a logical inference is that it sees his
accession as an accomplished fact.
52 . Davis, Regesta , xvi – xxi; V.H. Galbraith, ‘ Girard the Chancellor ’ , ante , xlvi (1931), 77 – 9.
53 . C. Johnson, ed., Hugh the Chantor: The History of the Church of York, 1066 – 1127 (London,
1961), 11; B. 278 and B. 352: Bates, Regesta , 835 – 6 and 1003 – 4.
54 . Historia Novorum II.68: G. Bosanquet, tr., Eadmer’s History of Recent Events in England.
Historia Novorum in Anglia (London, 1964), 71.
55 . DB Devon i.117a.
1433
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Close analysis of the other sources for the 1087 succession reveals,
however, some ambiguity. Symeon of Durham and William of
Malmesbury both hint at a need for speed: ‘ he gave the kingdom of
England to his son William … [who] hurried off to England … ( festinato
adiit ), ’
56
‘ William, before his father had fi nally expired, had sailed away
to England (
antequam plane pater expiraret Angliam enauigauerat )
thinking it more to the purpose to secure his own future interests than
to attend the burial of his father’s body. To this end, he was neither
dilatory nor sparing in the distribution of funds. ’
57
Orderic as usual
paints a fuller picture:
58
… the king, fearing that rebellion might suddenly break out in a realm as
far-fl ung as his, had a letter to secure the recognition of the new king
addressed to Archbishop Lanfranc and sealed with his seal. Giving it to his
son William Rufus, he ordered him to cross to England without delay. Then
he gave him his blessing with a kiss, and sent him post-haste ( properanter
direxit ) overseas to receive the crown.
There are also two curious small passages in William of Malmesbury:
Robert ‘ forfeiting both his father’s blessing and his inheritance, failed to
secure England’, and of Rufus, ‘ his hopes gradually rose and he began
to covet the succession’.
59
Orderic also reported King Malcolm of
Scotland in 1091 as saying ‘ I owe you nothing, King William [Rufus] …
but if I could see King William’s eldest son, Robert, I would be ready to
offer him whatever I owe ’ and to Curthose himself Malcolm said, ‘ King
William required my fealty to you as his fi rst-born son ’ , To which Robert
replied, ‘ what you allege is true. But conditions have changed and my
father’s decrees have been undermined in many ways. ’
60
Robert of Torigni seems to go further:
61
let me give you an account of his death, as some say it happened … [he]
granted the kingdom of England to his son William … [who crossed] to
England as swiftly as possible, where he was accepted … When urged to
reconquer by force the kingdom of England, taken away from him by his
brother, Robert is said to have answered with his usual simplicity and, if I
may put it so, almost as a fool: ‘ By the angels of God, if I were in Alexandria,
the English would have waited for me and they would never have dared to
make him king before my arrival. Even my brother William, whom you say
has dared to aspire to the kingship, would never risk his head without waiting
for my permission. ’
56 . Symeon of Durham, Historia Regum 169: T. Arnold, ed., Symeonis Monachi Opera Omnia.
II. II Historia Regem … (London, Rolls Series, lxxv, 1885), 214.
57 . Gesta Regum Anglorum III.283: eds. Mynors et al. , 512 – 13.
58 . Oderic Vitalis VII.16: ed. Chibnall, iv, 96 – 7.
59 . Gesta Regum Anglorum III.274 and IV.305: eds. Mynors et al. , 502 – 3, 542 – 3.
60 . Oderic Vitalis VIII.22: ed. Chibnall, iv, 268 – 71.
61 . Gesta Normannorum Ducum VII.44 and VIII.2: ed. van Houts, ii, 192 – 5 and 202 – 5.
EHR, cxxiii. 505 (Dec. 2008)
1434
THE DE OBITU WILLELMI
The most overt remarks are in Eadmer’s Historia Novorum . He chose to
begin his book by comparing the blessings of the reign of Edgar with
that of ‘ Ethelred, because he had grasped the throne by the shedding of
his brother’s blood … monstrous wrongs were done which every year
increased and grew worse and worse’. Then when describing the
succession in 1087:
62
[Rufus], intent on seizing the prize of the kingdom before his brother
Robert, found Lanfranc, without whose support he could not possibly attain
the throne, not altogether favourable to the fulfi llment of this his desire.
Accordingly, fearing that any delay in his consecration might result in the
loss of the dignity which he coveted, he began, both personally and indirectly
by all whom he could get to support him, to make promises to Lanfranc
…
These and other passages suggest that the succession of Rufus to England
in 1087 was not an entirely clear-cut matter. Can De obitu Willelmi shed
any light on this situation?
Previous research on De obitu Willelmi has paid great attention to the
donation of the regalia, pointing to a 1096 × 1098 charter in favour of St
Etienne Caen which exchanges them for property in England.
63
English
concluded that the Conqueror ‘ made no declaration about the succession
to England. The regalia … may have been those of the king of England
who was also duke of Normandy, to remain at Caen until a new king-
duke could legitimately claim them. ’
64
Sapir and Speet noted that De
obitu Willelmi might be mentioning the regalia in a metaphorical sense,
to indicate that Rufus was heir to the kingdom. But they favoured using
this passage to propose a date for De obitu Willelmi ‘ s composition in
1096 – 7, associated with the construction of the Conqueror’s tomb and
the handover of Normandy from Curthose to Rufus on the eve of the
crusade.
65
They observed that the Caen charter, like De obitu Willelmi ,
describes the Conqueror’s death; it is also witnessed by Bishop John of
62 . HN 3, 5 and 25: tr. Bosanquet, 3 – 5 and 26; M. Rule, ed., Eadmeri Historia Novorum in
Anglia, et opuscula duo de Vita Sancti Anselmi et quibusdam miraculis ejus (London, Rolls Series,
lxxxi, 1884), 25. This compares interestingly with his statement in the Vita Anselmi: ‘ When the
renowned William King of the English died, his son William inherited [ obtinuit ] the throne. ’
R.W. Southern, ed., Eadmer: The Life of St Anselm Archbishop of Canterbury. Eadmeri monachi
Cantuariensis . Vita Sancti Anselmi, archiepiscopi Cantuariensis (London, 1962), 63. The other
Canterbury source, the Acta Lanfranci, says ‘ … Lanfranc chose his son William as king, even as
his father had desired … ’ ‘ … fi lium eius Willelmum, sicut pater constituit, Lanfrancus in regem
elegit … ’ J.M. Bately, The Anglo-Saxon Chronicle. A Collaborative Edition. Volume iii. Manuscript A
(Cambridge, 1986), 87. This is in Hand 13 of the A manuscript, which was written after 1093. See
also the discussion in English ( William the Conqueror and the … Succession ), at 229 – 32; and J.S.
Beckerman, ‘ Succession in Normandy, 1087, and in England, 1066: The role of Testamentary
Custom, ’ Speculum , xlvii (1972), 258 – 60.
63 . The charter is printed in full in Sapir and Speet ( Kritische ), 56 – 7. Davis, Regesta , no. 397.
64 . English ( William the Conqueror and the … Succession ), at 236.
65 . Sapir and Speet, Kritische , 30 – 2.
1435
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Bath and Bishop Gerard of Hereford, who may be the John and Gerard
of De obitu Willelmi .
These apparent connections may, however, be coincidental. The
wording of the Caen charter concerning the regalia is very different
from that in De obitu Willelmi . Also, there are three other bishops and
ten laymen who witness the charter, besides John and Gerard, many of
them closely associated with Rufus’s rule in England. It is not clear
where the charter was attested, and there are suffi cient references in the
sources to both John and Gerard assisting Rufus in various ways for
their presence together as witnesses to be unremarkable. There is also a
more fundamental objection to linking the production of De obitu
Willelmi to this charter, namely that there is no obvious purpose that it
could have served in 1096 – 7. It was suggested by van Houts that Rufus
needed De obitu Willelmi to strengthen his claim to the regalia, and that
De obitu Willelmi was a ‘ piece of propaganda ’ written for that purpose;
she proposed Gerard as a possible author.
66
While De obitu Willelmi
does read like propaganda (but against Curthose, rather than in favour
of Rufus) it remains unclear who would have had reservations about the
release of this set of regalia to Rufus in 1096 – 7 and yet would have been
swayed by De obitu Willelmi . The possibility has to be considered
therefore that the question of the regalia has become a red herring,
hindering a fuller investigation of De obitu Willelmi ’ s signifi cance.
It is nevertheless apparent that the regalia are deliberately included in
De obitu Willelmi . The Vita Hludouuici model was modifi ed to stress
that Rufus was to be the recipient, not just that they are to be handed
into neutral care until the succession is decided,
67
and only after this
was the text cut. If Rufus’s succession, or the grant to him of some token
of royalty, were irrelevant, this sentence could have been omitted. So,
obliquely, De obitu Willelmi acknowledges Rufus as the next king by his
father’s consent, but this is not its main purpose.
There are fi ve features of the 1087 succession which merit particular
comment. First is Rufus’s hasty departure, even before his father had
died, at a time when there was no obvious external foe. The need for
speed is mentioned several times, and De obitu Willelmi does not actually
say he was present when his father died. This suggests a real fear, and
one obvious cause is lest a rival, presumably Curthose, beat him to the
throne. Orderic said that Rufus’s only companion on his journey was
Robert Bloet, who replaced Gerard as chancellor. It would have taken
two days to ride from Rouen to the port at Touques, longer if they
crossed from Wissant.
68
One can imagine them waiting with a boat
66 . Gesta Normannorum Ducum : ed. van Houts, i, lxiv – lxv.
67 . The imperial regalia were left with the widowed empress on several occasions from 1024
onwards, and Orderic Vitalis was aware of this convention. French kings sometimes left their
regalia to St Denis: M. Chibnall, The Empress Matilda. Queen Consort, Queen Mother and Lady of
the English (Oxford, 1991), 40 – 3.
68 . Oderic Vitalis VII.16 and X.2: ed. Chibnall, iv, 97 and v, 203.
EHR, cxxiii. 505 (Dec. 2008)
1436
THE DE OBITU WILLELMI
until they heard the king was dead, so as to minimise delay. Here, De
obitu Willelmi differs from the other accounts of the deathbed, by
implying that Rufus was present for the grant of the regalia. This seems
to be achieved deliberately, by the alteration of misit to permisit. This
may be a device to gloss over Rufus’s absence when the succession was
declared, or to stress that he was present at this critical time, only
departing subsequently. This lack of clarity is echoed in Stephen’s
accession; he was accused by the bishop of Angers, ‘ As for your statement
that the king changed his mind, it is proved false by those who were
present at the king’s death. Neither you nor Hugh could possibly know
his last requests, since neither was there. ’
69
Secondly, there is the absence of Curthose. He was heir to Normandy
and Maine, if no more. His father lay sick at Rouen for well over a
month, and if Curthose was indeed at Abbeville, as Robert of Torigni
says, he could have been at the bedside in a matter of days.
70
The
conventional wisdom is that by 1087 Curthose had been ‘ in rebellion ’
for several years, but a close examination of the contemporary sources
does not bear this out. On the contrary, this story can be shown to rely
on one passage in Orderic Vitalis, written when Curthose had already
been Henry I’s prisoner in England for many years, discredited and
disinherited. Orderic contradicts himself in another passage, saying that
Curthose had ‘ only recently ’ ( tunc nouiter ) left court, and the other
chroniclers are agreed that William the Conqueror was able to move a
large army to England in 1085 to counter the threatened Danish invasion,
and then he remained there himself until late summer 1086.
71
These
actions do not sit well with the idea of his son staging a major rebellion
at the time. It seems unlikely moreover that Curthose would have
imperilled his inheritance by deliberately staying away, and in the
absence of any description of a major rift with his father at this late
stage, except the contradictory comments of Orderic, the question arises
whether he might have been deliberately kept from knowing how ill his
father was.
Thirdly, Lanfranc was apparently unprepared for the king’s decision.
He was a close adviser and supporter of the Conqueror, yet Eadmer
stressed that Lanfranc had no idea that Rufus was the chosen heir.
Archbishop and king corresponded regularly, yet all the indications are
that no message was sent to Lanfranc between July and September 1087.
The only letter that is mentioned is the one Orderic said was carried by
69 . M. Chibnall, ed., John of Salisbury’s Memoirs of the Papal Court. Ioannis Saresberiensis
Historia Pontifi calis (London, 1956), 85; cited in W.C. Hollister, Henry I (New Haven and London,
2001), 479.
70 . Gesta Normannorum Ducum VIII.2: ed. van Houts, ii, 202 – 3.
71 . Oderic Vitalis VII.14 and V.10: ed. Chibnall, iv, 80 – 1 and iii, 112 – 13; ASC E for 1085 and
1086: D. Whitelock, D.C. Douglas and S.I. Tucker, eds., The Anglo-Saxon Chronicle. A Revised
Translation (London, 1961), 161 – 2; K. Lack, Conqueror’s Son: Duke Robert Curthose, Thwarted King
(Stroud, 2007), 27 – 35.
1437
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Rufus. This letter has not survived. Barlow observed that this letter only
occurs in the Norman tradition; ‘ neither the letter nor its tenor was
preserved at Canterbury. It would have been sensible for Lanfranc to fi le
this important mandate, but whoever collected and published his
correspondence omitted it, and Eadmer passes over it in silence. If
indeed such a letter ever existed … . ’
72
There was a two-week interval before Lanfranc crowned Rufus, during
which Eadmer claimed considerable pressure was applied. An additional
factor in Lanfranc’s deliberations may have been the knowledge that
both Harold and William the Conqueror had been crowned by the
archbishop of York, and in the delicate state of the primacy dispute,
Lanfranc would not want Rufus to apply to Archbishop Thomas.
Fourthly, one might expect some evidence that Rufus had been
publicly recognised as heir to the kingdom, or had at least supplanted
Curthose in his father’s charters. But there is no such evidence. The one
time when the Conqueror might have made some statement about the
succession, at Salisbury in 1086, it seems that he instead continued to
demand personal loyalty to himself alone.
73
No original charters place
Rufus before Curthose; two eighteenth-century copies of charters
apparently do so, but the latest two surviving acta have the normal
sequence of attestations. In the last two surviving acta he witnessed for
his father, Rufus was not even given the title comes , despite the fact that
one of them was witnessed in England.
74
Fifthly, there is the reaction of the magnates on both sides of the
Channel to Rufus’s succession. The story in the sources varies, and is
complicated by the hostility between Lanfranc and Odo, but the barons
moved quickly in support of the elder brother’s claim. The plot may
have been hatched in Normandy before Christmas; by Easter 1088 there
was a widespread uprising imminent in England. Interestingly, there
was no comparable rising against Curthose in Normandy. This rising
needs to be seen not merely in the context of the acknowledged status
of an anointed king and the oaths recently sworn to him,
75
but of what
the rebels stood to lose in practical terms. Although, as Strevett has
noted, Rufus was supported by many of the nobility based in England,
the three greatest lay landholders there and the Conqueror’s fi ve closest
supporters all rose for Curthose in 1088.
76
Odo had been imprisoned by
72 . F. Barlow, William Rufus (London, 1983), 55. Barlow points to a possible parallel with a letter
from Henry I to the pope, which Anselm omitted from his letter collection.
73 . ASC E for 1086: ed. Whitelock, 162.
74 . B. 205 (June 1082) and B. 279 (1083), B. 252 (January 1084) and B. 156 (probably Christmas
1085), B. 146 (April 1086 or later, ‘ fi lii regis Willelmus et Henricus ’ ) and B. 242 (late 1086 × 1087,
‘ fi lius regis ’ ): Bates, Regesta , 644 – 6, 837, 763, 513, 482 – 4, 741 – 2.
75 . M. Strickland, ‘ Against the Lord’s Anointed: Aspects of Warfare and Baronial Rebellion in
England and Normandy, 1075 – 1265 , in G. Garnett and J. Hudson, eds., Law and Government in
Medieval England. Essays in Honour of Sir James Holt (Cambridge, 1994), 56 – 79.
76 . N. Strevett, ‘ The Anglo-Norman Civil War of 1101 Reconsidered ’ , Anglo-Norman Studies ,
xxvi (2003), 159 – 75 at 160 – 1.
EHR, cxxiii. 505 (Dec. 2008)
1438
THE DE OBITU WILLELMI
the Conqueror, but the other magnates who sided with Curthose are
never known to have wavered in their loyalty. Of the rebel leaders, only
Eustace of Boulogne had relatively little to lose, since his main lands
were on the continent.
Donald and Bennett have both suggested that this rising represented
the magnates asserting their rights as Normans, not bound by Anglo-
Saxon conventions, pushing for a unitary succession and their own
newly achieved power as kingmakers, rather than personal loyalty to
Curthose.
77
An echo of this may survive in the speech that Orderic gave
to them:
78
If we serve Robert duke of Normandy as we ought we will offend his brother
William … Again, if we obey King William dutifully, Duke Robert will
confi scate our inherited estates in Normandy … since King William is the
younger of the two and very obstinate and we are under no obligation to
him he must be deposed or slain. Then let us make Duke Robert ruler over
England and Normandy to preserve the union of the two realms, for he is
older by birth and of a more tractable nature, and we have already sworn
fealty to him during the lifetime of the father of both men.
Whatever the underlying complexities, the 1087 succession certainly
occurred within a framework of varying expectations, and at a time of
transition. It conformed to Norman tradition for Normandy, but does
not seem to have followed English practices very closely, apparently
relying on designation, with a minimum of election and a move towards
‘ pre-emptive anointing’.
79
A few surviving hints in the sources, and
particularly the events of 1088, indicate that the succession of Rufus was
not the universally expected outcome, nor was it accepted without
contention.
Modern historiography has not always found the partition of 1087
comfortable either. Barlow and Le Patourel have stressed the inevitability
of divided loyalties when the Anglo-Norman lands were split, and these
were diffi culties that the Conqueror must have foreseen.
80
Bates and
others have suggested that a man so apparently eager to retain power
might wish to pass on his ‘ empire ’ intact, thereby enhancing his own
posterity. It is also certain that England was the richer of the two realms:
would William willingly deprive Normandy of this newly acquired
source of wealth? ‘ There is in fact a lot to be said for the emergence of
77 . M. Donald, King Stephen (London, 2002), 44. M. Bennett, ‘ Poetry as History? The ‘ Roman
de Rou ’ of Wace as a Source for the Norman Conquest’, Anglo-Norman Studies , v (1983), 21 – 39,
at 36.
78 . Oderic Vitalis VIII.2: ed. Chibnall, iv, 122 – 5.
79 . G. Garnett, ‘ Coronation and Propaganda: Some Implications of the Norman Claim to the
Throne of England in 1066 ’ , Transactions of the Royal Historical Society , xxxvi (1986), 91 – 116 at 93
and 115 – 6.
80 . Barlow ( William Rufus ), 40 – 5; J. Le Patourel, Feudal Empires, Norman and Plantagenet
(London, 1984).
1439
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
an awareness among the Anglo-Norman aristocracy and ruling family
of the importance of keeping Normandy and England united. ’
81
Strevett
has more recently argued that ‘ substantial sections of the aristocracy
clearly doubted whether the decision taken to divide Normandy and
England in 1087 was either legally correct or politically viable
’
,
82
although as Holt has noted, the term ‘ law ’ is scarcely applicable to this
period, except as an ‘ assemblage of customs and conventional practices
which were still malleable ’ .
83
The possibility that De obitu Willelmi could be part of the process of
promoting Rufus at the expense of Curthose’s more obvious claims to
both realms can now be assessed. Is it possible to narrow down the
dating limits for De obitu Willelmi ’ s composition? It is likely that it
predates 1100, principally because Henry I is omitted from it. Previous
attempts to date it more closely using the Caen charter and the donation
of the regalia are, however, not entirely convincing. The few words
devoted to the regalia in De obitu Willelmi may rather refl ect their
limited signifi cance at the time. Two other aspects of De obitu Willelmi
invite consideration as possible avenues for dating it. First, Eadmer said
that Anselm was present for part of the Conqueror’s illness, but he is not
mentioned in De obitu Willelmi .
84
Although an important abbot in
Normandy, Anselm only became signifi cant in England when he was
consecrated archbishop of Canterbury in 1093, so it may be that De
obitu Willelmi predates this. Secondly, and more positively, Robert of
Mortain was given a very prominent role in De obitu Willelmi , as a
magnate close to the Conqueror who approved his designation of Rufus.
But he was also a key player in the 1088 rising in favour of Curthose,
and died quite soon after, certainly by 1095.
85
On this basis, one might
ask if De obitu Willelmi could date to a time when Robert of Mortain
was still active and Anselm not yet archbishop.
Two questions need to be asked about De obitu Willelmi to explain its
presence in Gesta Normannorum Ducum . Why was it written? And why
did it survive in such a geographically distinctive manner? In answering
these questions, one possibility is that De obitu Willelmi has an earlier
origin than has been previously suggested, namely the 1088 rising. It could
be seen as propaganda for Rufus, but indirectly so, by discrediting Curthose
as a potential ruler. Its intended audience may have been a group who had
relatively limited experience of Curthose, for example the nobility based
in England. This fi ts with the simple allusion to Rufus as his father’s royal
heir, which was by then an accomplished fact in England.
81 . D. Bates, ‘ Normandy and England After 1066 ’ , ante , civ (1989), 851 – 80 at 872.
82 . Strevett, The Anglo-Norman Civil War … , 162.
83 . J.C. Holt, ‘ Politics and Property in Early Medieval England ’ , Past and Present , lvii (1972),
3 – 52 at 9.
84 . HN 23-25: tr. Bosanquet, 25.
85 . B. Golding, ‘ Robert of Mortain ’ , Anglo-Norman Studies , xiii (1990), 119 – 44 at 122; id.,
‘ Robert, Count of Mortain, magnate ’ , Oxford Dictionary of National Biography (Oxford, 2004).
EHR, cxxiii. 505 (Dec. 2008)
1440
THE DE OBITU WILLELMI
Another example of eleventh-century propaganda, the mid-century
work the Encomium Emmae Reginae , has historical inconsistencies and
a limited circulation (only one early manuscript survives) which invite
comparisons with De obitu Willelmi . There was no ‘ public opinion ’ that
could be swayed by pamphleteering in the eleventh century, but a
specifi c target audience, with limited access to alternative sources of
information, might be persuaded. On this basis, Lifshitz proposed that
Encomium Emmae Reginae was written in Flanders, while Emma was in
exile there, to persuade Baldwin V and his court to support Hardacnut’s
claim to the English throne.
86
John has called the Encomium Emmae
Reginae ‘ writing for contemporaries in a political crisis, ’
87
and Stafford
highlights several other features of it which are paralleled in the De obitu
Willelmi . Both seem to keep their benefi ciary out of the limelight; both
describe contemporary events so as to ‘ conjure … reality ’ ; both present
a specifi c explanation for unexpected political actions.
88
Early European biographical writing ‘ always embraced persuasion as
much as verisimilitude, ’ and the medieval forms of the genre were fi rmly
based on these foundations.
89
For the men most closely related to
Curthose, Henry I’s biographers have been scrutinised by Cooper, and
Bates has compared the standpoints of those who wrote for and about
William the Conqueror.
90
Specifi c political propaganda was also
certainly a well-established part of Anglo-Norman realpolitik. Most
notably, Harold was reinvented as ‘ count ’ not ‘ king ’ in the Conqueror’s
charters, and by 1086 there are only occasional hints that his reign had
ever occurred.
91
As Cowdrey has demonstrated, eleventh-century deathbeds, both
royal and ecclesiastical, were a focus for much falsifi cation, ranging
from pious interpretation to blatant forgery, but much of it within
contemporary notions of acceptability. Sometimes, these interpretations
86 . F. Lifshitz, ‘ The Encomium Emmae Reginae: A “ Political Pamphlet ” of the Eleventh
Century? ’ The Haskins Society Journal , i (1989), 39 – 50.
87 . E. John, ‘ The Encomium Emmae Reginae: A Riddle and a Solution’, Bulletin of the John
Rylands Library , lxiii (1980), 58 – 94 at 94.
88 . P. Stafford, Queen Emma and Queen Edith: Queenship and Women’s Power in Eleventh-
Century England (Oxford, 1997), 28 – 52.
89 . D. Bates, J. Crick and S. Hamilton, eds., Writing Medieval Biography, 750 – 1250: Essays in
Honour of Professor Frank Barlow (Woodbridge, 2006), 1.
90 . A. Cooper, ‘ “ The Feet of Those That Bark Shall Be Cut Off ” : Timorous Historians and the
Personality of Henry I ’ , Anglo-Norman Studies , xxiii (2001), 47 – 67; D. Bates, ‘ The Conqueror’s
Earliest Historians and the Writing of his Biography ’ , in Bates, Crick and Hamilton, eds., Writing
Medieval Biography … , 129 – 41.
91 . For example, B. 223 (1066 × 1067) refers to him as ‘ Harald kinge, ’ and B. 286 (May 1068)
calls him ‘ Haroldus vero rex, ’ but B. 300 (only datable to 1066 × 1075) has ‘ Edwardus rex ’ and
‘ Haroldus comes. ’ In B. 226 (1077 × 1080), the latest surviving charter to mention him, he is simply
‘ Harold ’ . Bates, Regesta , 710, 863 – 5, 896 – 7, 715 – 16. In Domesday he is almost always just ‘ Harold ’ ,
with an interlinear ‘ comes ’ where necessary to distinguish him from other men of the same name.
There are occasional exceptions such as for Soberton, Hants, which notes ‘ Harold took it while he
was reigning ’ ( ‘ quando regnabat ’ ) DB Hants i.38b; Garnett, ‘ Coronation and Propaganda ’ , 91 – 116,
especially at 99 – 109.
1441
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
(Cowdrey cites the Bayeux Tapestry here) were ‘ to give a propagandist
justifi cation to an interpretation of events at odds with what really
happened ’ but which had ‘ become current with the lapse of time ’ and
represented the divine purpose.
92
This is the context within which De
obitu Willelmi can be seen, as political propaganda associated with the
Conqueror’s death. Sharpe has, furthermore, noted that medieval
pamphlets ‘ are less likely to survive in their original, separate form than
when copied into a volume likely to be preserved in a library. This can
easily obscure their character. ’
93
Again, the parallels with De obitu
Willelmi are interesting.
If, then, one accepts that De obitu Willelmi could originally have been
a ‘ political pamphlet’, can anything be suggested about its authorship?
van Houts proposed Gerard as the author, but a problem with this
theory is that there is no obvious time in the careers of Rufus and Gerard
for its production. For John, however, the other man inserted in the De
obitu Willelmi text, there is such a time. It has already been seen that the
content of De obitu Willelmi is compatible with it having been written
in association with the 1088 rising, and it has been demonstrated that
John emerged from obscurity to being consecrated bishop of Wells in
July 1088. Allowing for a short interval between his nomination by
Rufus and the consecration, this could be construed as a reward for
performing some service to the king in the winter or spring 1087 – 8: just
the period when De obitu Willelmi would have been produced if it
played a part in the events of that year.
The only fi rm evidence that ties John to De obitu Willelmi is the
presence of his name in it. But there are some other pointers. The
‘ Historiola ’ of Wells describes him ‘ having been employed by the king
in many and great affairs, and having in consequence grown into
familiarity with him, he begged of the king for himself the city of Bath
… ’ .
94
Despite his otherwise rather unfavourable tone, William of
Malmesbury says John encouraged learning and improved the library at
Bath,
95
which suggests he was a man capable of composing such a piece.
Then there is his rapid promotion by Rufus, which is otherwise
unexplained. Soon after his consecration, Rufus also gave him Bath
Abbey, and confi rmed the award in 1091.
96
In this latter respect he is a
better candidate for the authorship than Gerard, who waited many
years for his reward.
92 . H.E.J. Cowdrey, ‘ Death-bed Testaments ’ , Monumenta Germaniae Historica, Schriften , Band
33, iv (1988), 703 – 24, especially at 723.
93 . R. Sharpe, ‘ Symeon as Pamphleteer ’ , in D. Rollason, ed., Symeon of Durham. Historian of
Durham and the North. Studies in North-Eastern History (Stamford, 1998), 214 – 29, at 214.
94 . J. Hunter, ed., A Brief History of the Bishoprick of Somerset from Its Foundation to the Year
1174 . Camden Society, viii, 1840, 21 – 2. The ‘ Historiola ’ is a late twelfth-century production.
95 . Hamilton ( William of Malmesbury’s Gesta Pontifi cum ), II.90: 194; Gesta Regum Anglorum .
IV.341: eds. Mynors et al. , 591.
96 . Hunt ( Two Chartularies … of Bath ), 40 – 2.
EHR, cxxiii. 505 (Dec. 2008)
1442
THE DE OBITU WILLELMI
The Vita Hludouuici , from which the ‘ propaganda ’ material is drawn,
seems to have been relatively scarce in England compared with France.
Only two manuscripts with possible English connections survive,
97
and
this combined with the fact that the closest surviving version to that in
De obitu Willelmi originated in the Chartres area, points to a north-
French or Norman origin for De obitu Willelmi . This is also compatible
with John’s authorship, since he is absent from the English records until
his consecration. One could thus imagine John using material closely
related to the Chartres manuscripts, to assemble De obitu Willelmi from
the Vita Hludouuici and the Vita Karoli Magni , on the instructions of
someone in Rufus’s entourage, as soon as rumours of the impending
rising began to circulate.
Rufus is described using every means available to win people over to
his side in early 1088.
98
A document such as De obitu Willelmi could
have been a powerful weapon, if used with discretion to sway essentially
illiterate audiences among the nobility, and a memory of this may even
have provided the story reworked by Orderic to portray Rufus arriving
in England with a letter of recommendation from his father. This date
for the composition of De obitu Willelmi is compatible with the names
of people included and omitted, explains the heavy emphasis on
Curthose’s failings and accounts for the rapid promotion of John in mid
1088. It also fi ts with the story of the Conqueror reluctantly dividing his
lands, wishing to disinherit Curthose completely, but being tied by his
previous promise to grant him Normandy. The implication is that Rufus
would otherwise have inherited everything.
Supposing that De obitu Willelmi was produced for Rufus, probably
on the continent, and then used to infl uence the nobility in England to
abandon Curthose’s cause in 1088, is it possible to construct a plausible
history for it thereafter? Here one is inevitably moving further into the
realms of speculation, but such a history must at least be possible, or the
fi rst part of the hypothesis will fail. Some sequence of events must have
caused the B redaction of
Gesta Normannorum Ducum
, which is
favourable to the dukes of Normandy but ambivalent about Curthose,
to be combined with De obitu Willelmi which is very hostile to him.
Two separate models can be explored briefl y here. Either the B redaction
of Gesta Normannorum Ducum could have survived intact at Durham
by chance, in which case no special link with Durham need be sought,
or alternatively the Durham B2 manuscript (the oldest and only
complete copy) could be seen as a key to the formation of the B
redaction.
97 . Vatican Apostolic Library, Reginensis lat.692, and Oxford Bodley 755, both of which
date to the second half of the twelfth century: von Trempe ( Theganus Gesta Hludowici ), 33 – 4 and
123 – 33.
98 . For example, JW for 1088: ed. McGurk, iii, 50 – 1.
1443
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Propaganda such as De obitu Willelmi would probably have been kept
initially among Rufus’s royal documents. Three men were associated
closely with Rufus’s administration from early in his reign: Gerard,
Robert Bloet who Orderic says accompanied Rufus on his dash for the
throne and who replaced Gerard as chancellor, and Ranulf Flambard,
later appointed bishop of Durham. Any of these could have preserved
the original copy of De obitu Willelmi .
Gullick concluded that the B2 manuscript is so different from other
Durham books that it is unlikely to have been made there. He identifi ed
Symeon of Durham’s hand in the fi rst eight lines, as well as the
annotations. The two scribes who worked on it have continental hands
and Gullick therefore proposed that it was created in Normandy or
France, under Symeon’s supervision.
99
There are two obvious
opportunities for contact between Normandy and Durham: Bishops
William of St Calais (1088 – 91) and Ranulf Flambard (1101) both spent
time in exile in Normandy. Symeon himself is thought to have moved
to Durham in the early 1090s, after Bishop William’s exile.
If then the B2 manuscript survived at Durham simply because the
Durham library was fortunate in its later history, one could imagine a
new version of Gesta Normannorum Ducum being created for Rufus in
Normandy soon after 1096, and De obitu Willelmi being added to it as
a suitably fl attering end. This would account for the B redaction’s muted
attitude to Curthose, who was still nominally duke, and the addition of
the independent material about earlier dukes, who were Rufus’s ancestors
as much as Curthose’s. Symeon’s interest in historiography, and William
of St Calais ’ reinvigoration of the Durham library, could readily explain
why a copy went to Durham, either a continental copy supervised by
Symeon or one made by continental monks who, like him, had moved
to England.
100
Alternatively, if Durham is pivotal in the creation of the B redaction,
one could hypothesise that
De obitu Willelmi
found its way
independently to Durham, perhaps with Ranulf Flambard when he
became bishop in 1099. By then, Symeon was beginning his career as a
writer rather than merely a copyist. It was possibly he who incorporated
Archbishop Thomas’s epitaph to the Conqueror into De obitu Willelmi ,
using an English copy of the text. This could explain the differences
between the De obitu Willelmi version of the epitaph and that in Orderic.
Either then or when making the fi nal copy of De obitu Willelmi which
was added to Gesta Normannorum Ducum , the phrase ‘ as later events
proved to be true ’ might also have been added. Flambard’s diffi culties in
1101, followed by Curthose’s defeat by Henry I, temporarily in 1101 and
99 . Gullick ( The Hand of Symeon of Durham , in Rollason , Symeon of Durham ).
100 . R. Gameson, ‘ English Book Collections in the Late Eleventh Century and Early Twelfth
Century: Symeon’s Durham and its Context ’ , in Rollason, Symeon of Durham , 230 – 53.
EHR, cxxiii. 505 (Dec. 2008)
1444
THE DE OBITU WILLELMI
defi nitively in 1106, provide two plausible occasions for the fi nal
production of B2, with De obitu Willelmi again supplying a politically
acceptable ending, at a time when the bishop of Durham was especially
eager to display his loyalty to the king. Of the two models for the
formation of the B redaction, this is perhaps the more robust.
After the Battle of Tinchebrai in 1106, Henry I wrote to Anselm,
announcing his victory as a triumph of God’s cause. Chibnall has
commented
101
… we may well ask ourselves if he took further steps to put out this statement,
which fi nds its way into several chronicles … Whether the king trusted to
his court circle and leading bishops to spread news for him, or deliberately
circulated his own interpretation of events … there can be no doubt that he
excelled in public relations. William Rufus, on the other hand, seems to
have been careless in such matters.
It has been the object of this study to suggest that De obitu Willelmi
could be one successful example of Rufus’s public relations machine
that has hitherto gone unnoticed.
University of Birmingham
KATHERINE LACK
101 . Oderic Vitalis , ed. Chibnall, i, 88.
1445
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
Appendix
1:
Parallel texts of De obitu Willelmi, Vita Hludouuici and Vita Karoli Magni
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
De obitu Willelmi, ducis Normannorum regisque
Anglorum, qui sanctam ecclesiam in pace uiuere
fecit Anno Dominie Incarnationis millesimo
octogesimo septimo piisime recordationis rex Wil-
lelmus, dum a Medante subuersione seu com-
bustione reuerteretur
cepit fastidio tabescere et
nausianti stomacho cibum potumque reicere,
crebris suspiriis urgere, singultibus quati ac
per hoc uirtute destitui.
From
chapter
62
–
64
.
Coepit ergo fastidio tabescere et nausianti
stomacho cybum potumque intendere, crebis
suspiriis urgueri, ingultibus quati ac per
hoc virtute destitui
. Natura enim suis deserta
comitibus, necesse est ut victa fatiscat.
From Chapter
30
–
33
Quod cernens iussit sibi parari habitaculum
apud ecclesiam sancti Geruasii, que est sita in
suburbio urbis Rotomagensis,
ibique uiribus
desertus lecto sese committit.
Quod cernens, iussit sibi parari habitacula
aestiva atque expeditionalia in insula quadam
contigua Mogontiace civitati,
ibique viribus
desertus lectulo sese committit.
Porro quis explicet pro eccelesie statu sollici-
tudinem, uel pro eius concussione merorem?
Porro quis eius explicet por ecclesie statu
sollicitudinem, vel pro eius concussione
merorem?
EHR, cxxiii. 505 (Dec. 2008)
1446
THE DE OBITU WILLELMI
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
Quis narrare lacrimarum fl
umina quas pro accel-
eratione diuine fundebat clementie? Non enim se
recessurum dolebat, sed quod futurum nouerat
gemebat
affi
rmans Normanniam patriam esse post
suum obitum miseram, sicut postea rei probauit
euentus.
Quis narrare lacrimarum fl
umina, quae pro
acceleratione divine fundebat clementiae?
Non enim se recessurum dolebat, sed quod
futurum noverat gemebat
, dicens se miserum,
cuius extrema talibus clauderentur miseriis.
Aderant autem eius consolationi uenerabiles antis-
tites et alii serui Dei plurimi inter quos erant
Aderat autem eius consolationi venerabiles
antistites et alii servi Dei quamplurimi; inter
quos erant
Willelmus archiepiscopus prefate urbis, Gislebertus
episcopus Lexouiensis, Iohannes medicus et Gerardus
cancellarius,
Heti venerabilis Treuerorum archiepiscopus,
Otgarius Mogontiae similiter archiepiscopus
sed et Robertus comes Moritoniensis, frater eiusdem
regis,
sed et Drogo frater domini imperatoris Meten-
sis episcopus necnon sacri palatii archicapel-
lanus,
quem quanto sibi propinquiorem nouerat, tanto
ei familiarius sua omnia credebat.
quem quanto sibi propinquiorem nouerat,
tanto ei familiarius sua omnia
et semet
credebat.
Per eum quidem cotidie confessionis sue munus
sacrifi
tiumque et cordis humiliati, quod Deus
non despicit, offerebat. Cybus eius erat solum-
modo per XL dies dominicum corpus, laudante eo
iustiam Dei et dicente
‘ Justus es, Domine, ut quia
quadragesime tempus non ieiunans exegi saltem,
coactus idem tibi ieiunium exsolvam.
’
1447
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
Iussit autem eidem uenerabili fratri suo
Roberto
ut
ministros camere sue ante se uenire faceret et
rem familiarem que constabat in
thesauris
regali-
bus scilicet coronis, armis, uasis, libris uestibusque
sacerdotalibus, per singula describi iuberet. Et
prout sibi uisum fuit, quid ecclesiis, quid pauperi-
bus, postremo quid fi
liis largire deberet edixit.
Iussit autem eidem venerabili fratri suo
Drogoni
ut
ministros camerae suae antevenire faceret
et rem familiarem que constabat in
orna-
mentis
regalibus scilicet coronis, armis, vasis,
libris vestibusque
sacerdotalibus, per singula
describi iuberet. Et prout sibi uisum fuit,
quid ecclesiis, quid pauperibus, postremo
quid fi
liis largiri deberet, edixerat,
Hlothario
scilicet et Karolo.
VKM
From
chapter
33
- Charlemagne
’ s will
thesauris suis atque pecunia, quae in
illa die in camera eius inventa est
…
quae in auro et argento gemmisque et
ornatu regio in illa ut dictum est, die
in camera eius poterat invenire omnia
ex aere et ferro aliisque metallis vasa
atque utensilia cum armis et vestibus
.. alioque aut pretioso aut vili
…
Et Willelmo
quidem suo fi
lio coronam, ensem,
sceptrum gemmisque redimitum habendum
permisit.
Et
Hlotharo
quidem suo fi
lio coronam, ensem,
sceptrum gemmisque redimitum
eo tenore
habendum
misit, ut fi
dem Karolo et Iudith ser-
varet et portionem regni totam illi consentiret et
tueretur,
quam Deo teste et proceribus palatii
ille secum et ante se largitus ei fuerat
. [
see page
below
] His rite peractis, gratias Deo egit, quia
nichil sibi superesse proprium cognovit.
Inter hec tam uenerabilis antistes
Willelmus
quam
ceteri
qui aderant
,
Inter hec tam venerabilis antistes
Drogo
quam ceteri
pontifi
ces, dum in cunctus quae
agebantur Deo grates persolverent - utpote quia
videbant, eum quem chorus virtutum semper
comitatus fuerat, nunc persverantia subsequens,
quasi cauda hostiae, totum eius sacrifi
tium
vita
prorsus acceptum reddebat - unum erat, quo
gaudium eorum obfuscabatur :
EHR, cxxiii. 505 (Dec. 2008)
1448
THE DE OBITU WILLELMI
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
uerebantur
ne forte suo fi
lio primogenito Roberto
implacabilis esse uellet, scientes quod uulnus
requenter inscisum aut cautherio adustum
acerbiorem sustinenti propagaret dolorem; fi
si
tamen de eius inuicta pacientia, qua semper
usus est, per
archiepiscopum Willelmum,
cuius
uerba spernere nolebat, animum illius leniter
pulsant. Qui primum quidem amaritudinem sui
monstrauit animi.
At uero parumper deliberans
et uiribus quantuliscumquecollectis, enumerare
uidebatur quot et quantis ab eo affl
ictus sit inco-
modis, dicens:
verebantur
enim ne forte fi
lio
Hludouuico
implacabilis esse vellet, scientes quod vulnus
frequenter inscisum aut cautherio adhustum
acerbiorem sustinenti propagaret dolorem;
fi
si tamen de eius invicta patientia, qua sem-
per usus est, per
Drogonem fratrem eius,
cuius
verba spernere nolebat, animum illius leniter
pulsant. Qui primum quidem amaritudinem
sui monstravit animi,
at vero parumper delib-
erans et viribus quantuliscumque collectis,
enumerare conabatur, quot et quantis ab eo
affl
ictus sit incomodis
et quid contra naturam
et Domini praeceptum talia agendo commeru-
erit .
‘ Quia ispe,
’ inquit,
‘ uenire satisfacturus
non uult
‘ Sed quia ispe,
’ inquit,
‘ ad me venire satisfac-
turus
nequit,
aut
dedignatur
, ego quod meum est, ago:
ego quod meum est ago:
uobis testibus et Deo, omnia que in me peccauit,
illi remitto,
et omnem ducatum Normannie sibi concedo
’
(quem Deo teste et proceribus palacii illi
iam-
dudum
ante largitus fuerat).
[inserted from page
above]
uobis testibus et Deo, omnia que in me
peccauit, illi remitto.
1449
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
‘ Vestrum autem erit illum monere, ut, si ego illi
tociens perpere gesta indulsi, ille tamen sui non
obliuiscatur, qui canos paternos deducit cum
dolore ad mortem, et in talibus communis patris
Dei precepta minasque contempsit.
’ His dictis
petiuit ut in se celebraretur uisitacio et unctio
infi
rmorum,
et per offi
cium
archpresulis
et
per
manus eius iuxta morem communio sacra
sibi traderetur.
‘ Vestrum autem erit illum monere, ut, si ego illi
tociens perperam gesta indulsi, ille tamen sui
non obliuiscatur, qui canos paternos deducit
cum dolore ad mortem, et in talibus commu-
nis patris Dei precepta minasque contempsit.
’
His
peractis et
dictis
erat enim vespere sab-
bati - praecepit, ut ante se celebrarentur vigiliae
nocturne et ligno sanctae crucis pectus suum
muniretur; etquandiu valebat, manu propria tam
frontem quam pectus eodem signaculo insignibat,
si quando autem lassabatur, per manus fratris sui
Drogonis nutu id fi
eri poscebat. Mansit ergo tota
illa nocte omnis virtutis corporeae inops, solius
sobrietatis animi compos. In crastinum, que erat
dominica, iussit ministerium altaris praeparari
et per offi
tium
Drogonis missarum
offi
tia
celebrari
, necnon
per manus eius iuxta morem
communionem sacram sibi tradi
et post hec cuiusdam potiuncule calidule
haustum praeberi. Quo perpaululum praeli-
bato, precatus est fratrem et simul adstantes, ut
curandis operam darent corporibus, se tandiu
prestolaturum, quamdiu illi refi
ci
possent.
Instante autem migrationis eius articulo,
iunctis pollice cum articulis - hoc enim facere
consueverat, si quando fratrem nutu vocabat -
Drogonem
accersibat … .
EHR, cxxiii. 505 (Dec. 2008)
1450
THE DE OBITU WILLELMI
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
In talibus ergo uite presentis terminum sortitus
ad requiem feliciter, ut credimus, commigrauit
.
In talibus ergo vite praesentis terminum
sortitus ad requiem feliciter, ut credimus,
commigravit
, quia veraciter dictum est a
veridico doctore :
‘ Non potest male mori, qui
bene vixerit.
’
Decessit autem
quarto idus Septembris,
anno uite
sue
quinquagesimo nono,
et
Anglie
quidem prefuit
per annos
uiginti duos.
Decessit autem
XII kalendas iulii
anno vite
suae
LXIIII;
et
Aquitanie
quidem praefuit per
annos
XXXVII, imperator autem
XXVII.
From
chapter
9
Fuit autem ipse
rex omnium, qui sua etate gentibus
dominabantur, et prudentia maximus et animi
magnitudine prestantissimus, nichil in iis,
queuel
-
suscipienda erant uel exequenda, aut propter
laborem detrectauit, aut propter periculum
exhorruit, uerum unumquodque secundum suam
qualitatem et subire et ferre doctus, nec in
aduersis cedere, nec in prosperis falso blandienti
fortune assentire solebat.
Nam
rex, omnium qui sua aetate
gentibus dominabantur, et pruden-
tia maximus et animi magnitudine
prestantissimus, nihil in his,
quae
vel suscipienda erant vel exequenda
aut propter laborem detractauit,
aut propter periculum exhorruit,
verum unumquodque secundum
suam qualitatem et subire et ferre
doctus, nec in adversis cedere, nec
in prosperis falso blandienti fortu-
nae assentiri solebat.
1451
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
Corpore fuit amplo atque robusto, statura emi-
nenti, quae tamen iustam non excederet
.
From
chapter
22
Corpore fuit amplo atque robus-
to, statura eminenti, que tamen
iustam non excederet
nam septem suorem pedum pro-
ceritatem eius constat habuisse
mensuram apice capitis rotundo,
oculis praegrandibus ac vegitis, naso
paululum mediocritatem excedenti
…
EHR, cxxiii. 505 (Dec. 2008)
1452
THE DE OBITU WILLELMI
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
In cibo et potu temperatus, sed in potu temperan-
tior, quippe qui ebrietatem in qualicumque hom-
ine nedum in se ac suis plurimum abhominabatur.
From
chapter
24
In cibo et potu temperatus, sed
in potu temperantior, quippe qui
ebrietatem in qualicumque hom-
ine nedum in se ac suis plurimum
abhominabatur.
Cibo enim non. adeo abstinere
puterat, ut saepe quereretur noxia
corpori suo esse ieiunia. Conviva-
batur rarissime, et hoc praecipuis
tantum festivitatibus, tunc tamen
cum mango hominum numero.
Caena cotidiana quaternis tantum
ferculis praebebatur, praeter assam,
quam venatores veribus inferre
solebant, qua ille libentius quam
ullo alio cibo vescebatur. Inter
caenandum aut aliquod acroama aut
lectorem audiebat. Legebantur ei
historiae et antiquorum res gestae.
Delectabatur et libris sancti Augus-
tini, praecipueque his qui de civitate
Dei praetitulati sunt.
Vini et omnis potus
Vini et omnis potus
1453
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
adeo parcus in bibendo erat, ut post cenam raro
plus quam ter biberet.
adeo parcus in bibendo erat, ut
super caenam raro plus quam ter
biberet.
Aestate post cibum meridianum
pomorum aliquid sumens ac semel
bibens
…
Erat eloquencia copiosus et exuberans, poteratque
From
chapter
25
Erat eloquentia copiosus et exuber-
ans, poteratque
quicquid uellet apertissime exprimere,
quicquid vellet apertissime exprim-
ere,
Nec patrio tantum sermone
contentus …
uoce
rauca
quidem,
From
chapter
22
voce
clara
quidem,
sed que minus forme conueniret.
sed quae minus
corporis
formae
conveniret.
Religionem Christianum, qua ab infantia fuerat
imbutus, sanctissime et cum summa pietate coluit.
From
chapter
26
Religionem Christianum, qua ab
infantia
fuerat inbutus, sanctis-
sime et cum summa
pietate coluit,
ac propter hoc plurimae pulchritu-
dinis basilicam Aquisgrani extruxit
auroque et argento et luminaribus
atque ex aere solido cancellis et
ianuis adornavit
…
EHR, cxxiii. 505 (Dec. 2008)
1454
THE DE OBITU WILLELMI
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
Ecclesiam mane et uespere et sacrifi
cii
tempore,
quoad eum ualitudo permisit, impingre
frequentauit.
Ecclesiam et mane et vespere,
item nocturnis horis
et sacrifi
cii
tempore, quoad eum valitudo
permiserat, impingre
frequentauit.
From
chapter
31
corpus more sollemni lotum et cura-
tum et maximo totius populi luctu
ecclesiae inlatum atque humatum
est. Dubitatem est primo, ubi reponi
deberet, eo quod ipse vivus de hoc
nihil praecepisset.
Tandem omnium animis sedit nusquam eum hon-
estius tumulari posse, quam in ea basilica, quam
ipse
ob amorem et honorem Dei
et sancti Stephani
prothomartiris
Tandem omnium animis sedit
nusquam eum honestiustumulari
posse, quam in ea basilica, quam
ipse
obamorem et honorem Dei
et
Domini nostri Iesu Christi et ob
honorem sanctae et aeternae virginis,
gentricis eius,
proprio sumptu
in Cadomo
construxerat,
et sicut
antea disposuerat.
proprio sumptu
in eodem vico
construxit.
1455
EHR, cxxiii. 505 (Dec. 2008)
PROPAGANDA FOR THE ANGLO-NORMAN SUCCESSION
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
In hac
ergo
sepultus est
, et
arca argentea deaurata
supra tumulum eius est extructa
per fi
lium
suum
Willelmum, qui ei in regno successit Anglico, et
titulus in eadem huiusmodo aureis litteris scriptus:
In hac sepultus est
eadem die, qua
defunctus est,
arcusque supra tumulum
deauratus
cum imagine et titulo
extructus.
Titulus ille
hoc modo descriptus est:
SUB HOC CONDITORIO
SITUM EST CORPUS. KAROLI
MAGNI ATQUE ORTHODOXI
IMPERATORIS, QUI REGNUM
FRANCORUM NOBILITER
AMPLIAVIT ET PER ANNOS
XLVII FELICITER REXIT.
DECESSIT SEPTUAGENARIUS
ANNO DOMINI DCCCXIIII,
INDICTIONE VII, V. KAL. FEBR.
EHR, cxxiii. 505 (Dec. 2008)
1456
THE DE OBITU WILLELMI
De Obitu Willelmi
a
Vita Hludowici
b
Einhard’s Vita Karoli Magni
c
William the Conqueror’s epitaph in DOW
Qui rexit rigidos Normannos atque Britannos
Armis deuicit
fortiter optinuit
Et Cenomannenses uirtute cohercuit enses
Imperiique sui legibus applicuit
Rex magna parua iacet
tumulatus
in urna
Suffi
cit et magno parua domus domino.
Addiderat septem ter quinis Scorpius unam
Virginis in gremiis Phebus et hic obiit.
William the Conqueror’s epitaph in Orderic
Vitalis
d
Qui rexit rigidos Normannos atque Britannos
Audacter uicit
fortiter optinuit
Et Cenomannenses uirtute cohercuit enses
Imperiique sui legibus applicuit
Rex magna parua iacet
hac Guillelmus
in urna
Suffi
cit et magno parua domus domino.
Per septem gradibus se uolerat atque duobus
Virginis in gremiis Phebus et hic obiit.
a
Text taken from GND ed, van Houts, volume ii, pp.
184
–
90
. The complete text of DOW is here compared with the relevant sections of its two models.
b
Text
based
on
Engels
( De obitu Willelmi ducis
),
pp.
223
–
30
, with addition of
‘ sceptrum
’ from
the
P
1 mss. See von Tremp (
Uberlieferung
).
c
Text
taken
from
Engels
( De obitu Willelmi ducis
),
pp.
223
–
30
.
d
OV
VIII.
1: ed. Chibnall, volume iv pp.
110
–
13
.