Seigneurial Residence in Normandy

8/12/2019 Seigneurial Residence in Normandy

1/29

The Seigneurial Residence in ormandy, 25- 2 25: an Anglo- orman Tradition?

By EDWARD IMPEYENGLAND and Nonnandy skared a common /though nol exclu.rilJl) tradition in the design ofseigneurial houses between c. 25and J 225, typified y the housing ofthe hall and chambtrS inseparate buildings, both built to an increasingry standardized palltrn. The traJj/inn as known inEngland is briefly defined and a sekction 0/ the Nonnan evidence presented and discussed in thelight qf it, identifYing a common evolution during the 12th century but some differences in detail.It is then suggested that the pattern may have been particular to England and NonnandyJ and thatit originated in an AngLo-Saxon tradition transplanted to Normandy qfler the Conquest. TheNorman impad on domestic building in England is also briefly considered, with particularrtftrtnce to the Continental storeyed house and its best-known manifeslo.tion. tJu residential toWtr.

Several buildings indicating that the 12thcentury seigneury of England andNormandy shared at least one tradition in the design of their houses were presentedin a n arti cl e of 1993- Since then much new evidence has come to light, and whatfollows is an attempt to describe the nature and development of the shared traditionwhich can now be identified, and to consider two questions it raises: was thetradition, strictly defined, peculiar to England and Normandy?; and if so, wheredid it originate?THE SEIGNEURIAL RESIDE CE IN NORMAi DY c. . 5 -1 5 Fig. I)

THE HALL AND CHAMBER-BLOCK MODEL has b ee n accepte d for some time t ha t English buildings of the type oncecalled first-floor halls and ground-Aoor or end-bay halls were not alternatives,but routinely existed side by side, respectively housing the more private and morepublic quarters of the house. [t has also been shown that by c. 1200, bot h therelative positioning of the two elements and their particular form were becoming

increasingly standardized: the hall in the siting of its doorways, and the secondE Impcy, Sdgno:uriaJ domesl;C archilectUn: in Nonnandy, 1050-1350 ,82-120 in G. Mdrion:JorM:1 andM. Jones cds), Mrvwrial IJo Jlic BNilJm s r ,lmul nd .Jlurllw. tu t London, 9 13) 19 13 1)J. Since thenlOme or the rnalerial prac:nlcd hen: ha l been published in t n:nch E. Impcy, La Demeure IielgtlCuriaie enNonnandie entn: 1125 el 12250:1 tndition Anglo-Nonnandc , 219-4-2 in M. Baylt and P. Rouel cds),L AnltikctlUelUmfltllflil:a 2 \W Caen, 1997).

45


2/29

E D WA R D IM PE Y

_4 (

FIG. INormandy. Map howing location or ite mentioned in th e lUI.

building which can for convenience be called a chamber-block by (among otherfeatures) the raising of its inhabited level over a basement.2 As the hall an dchamber-block fonnula was not, of course, followed by llseigneurial builders ofthe period, the model must not be used to interpret every fragmentary remnant of12thcentury domestic architecture. No r must it be assumed that the functionaldistinction between hall an d chamber was in practice as rigid as the structuraldifferences between typical examples might suggest; indeed some interchangeability, or degree of private versus public use is actually implied by the looseapplication of the words aula an d camera in the Middle Ages.s But given thesereservations, an d although not universally acceptcd,4 the model offers the mostconvincing interpretation of many 2th- an d 13th-century English sites an dI \Y.J. Blair, Hall an d chamber: EnIiish domalK:p1anning 1000-12.)0 , 1-21 in G. Mdrion jonesand ~ J o n d(cds), A t w lD rw lK iltJiJrtsitt , Luaiad ofwtAml Ftaa London 993). On this WI point seeJ. Grenville:, Mffiittwl (uK:ester, 1997), 86 88. The principal crilie of the model hu bttn M. W. Thompson (see esp. Mffiittwll UMps }ff Jl J itt f I /_NIJVaIG (Akkrshot, 1998), 125 an d Tl ) e I). For further evidence an d dOCu ion of the model E. Impey an dR Ha/Ti , Boothby Pagncll revisilcd , in G. Mcirion:Jones el aI. (cds), 71w Seiptmal Rai nru itt (forthcoming); also lhe measured ~ n oflhe modc:l b) Grenville (op. cit. in nOle 3, 66-78).


3/29

S EI G NE U RI AL R ES ID E NC E IN N O R M A N D Y 47structures. As such, it is a valid basis for the in terpret ation of the structurallyanalogous buildings an d complexes which have now been identified in Normandy.T H E H A L L A N D C H A M B E R - B L O C K IN N O R M A N D Y

To date, fifteen Norman sites ar e known to retain at least fragments indicativeof a hall an d chamber-block arrangement: although no t abundant whencompared to those typical even of the period 1225-1350, still less those of14-50-1550) they cover a sufficient chronological, geographic and social range toimply the dominance of the type in this period.O f the most significant group - the four sites where elements of bothbuildings survive - the earliest is that at the chateau of Beaumont-Ie-Richard Calvados), where the remainsof a hall and a near-complete chamber-block, bothof c. 1150, stand no more than 10 m apart Figs. 2-4).5 O f the hall, the mostsubstantial fragments are the N. gable and an adjoining stretch of the western sidewall, the latter rising to about 6 m ab ov e the original floor level an d pierced by alarge window. Parallel an d opposite to this, the existing external wall incorporatestwo cylindrical piers, complete with scalloped capitals, shown by the springing ofan arch on the same axis to have carried an arcade. ca n b e d ed uc ed t ha t this wasof at least four bays, although the loss of the SW. gable wall leaves the possibilitythat there were one or two more. The building thus consisted, in essence, of a mainpart or n av e ) 8.5 m wide adjoined by a single aisle facing the bailey.

The second building, known as La Chapelle, was modified to serve as such inabout 1630, but the lack of an y earlier ecclesiastical features, its north-southorientation an d overall conformity to the typical chamber-block form leaves littledoubt as to its original function. Better preserved than thc hall, for present purposesits structural analysis needs no more explanation than offered in figure 4, althoughit might be pointed ou t t ha t the s ou th er n e nd of the internal blind arcading on theupper floor, coinciding with the t h ~ e n t u r y cross-wall, shows that there was anoriginal partition in the same position. The decorative detail, both inside and out,is among the more lavish at any building of its type in England or Normandy.At Creully, in the Bessin, 12 km E. of Bayeux, the single-aisled hall an d thegigantic chamber-block adjoining its W. side extend along the western edge of thecastle bailey, on a scarp overlooking the Seulles valley. Architectural detail suggestst ha t the complex dates from C.I r60-1 170, an d can thus be a tt ri bu te d to RichardofCreully, son of Robert of Gloucester.6 No single part of the ensemble is as perfectas La ChapelJe a t B ea umo nt , bu t the buildings are larger an d grander, while theunusual nature of the 14th- and r 7t h-c en tu ry modifications has allowed thepreservation of features routinely destroyed elsewhere.From the exterior, Romanesque fabric ca n o nl y be identified o n t he S. front Fig. 6 bu t here the profile of the entire elevation in its original form is clear

) For fuller description and additional drawings see E. Impey, op. cit. in note 1,85-94 an d E. mpey, L ArcluttclurtStignrorialt t Basse Jformand/t /5 /35 Rappon de Prospection Thcmalique, Saison, 1993 [ 993 2)]. Reportprepared for the Ministtre de la Culture, ~ 1 2 3 46 For historical notes and a fuller description SI: : E. Impey, Chlileau ik Creuf y Cabourg, 1995).


4/29

E W R IMPEY

Metres 1 ~ ~ ~ 1 - 1 >\ \ ~\ \ \

/ /

1 /

/

\ i\ \\ \-

FIG 2Beaumont Ie Richard Site plan showing the remains of the hall B) and the chamber block A). The outline ofthe defences is suggested by sections oflater wall and earthworks. urveyed nd dr wn A I 992


5/29

S E IG N E U RI A L R E SI D EN C E IN N O R M A N D Y 49

+ +

Inferred12lh century -3lh century I15th ccnlury -6th century -7th cenlury d 7 0 0 E J19th cenlury c Metres

FIO3Beaumollt-Ie-Richard. Croundf1oor plan suroq d and drawn A. I. M. A. I. J.-P. C. 1991 and sc


6/29

5 EDW AR D IM PEY

2th century h century th centuryr::::::J c 1630 19th century = Modern POsl medle\al _ .....1 0 5 Metres IG4Beaumont Ie Richard Plans an d section of the chamb


7/29

SE IGNEURIAL RESIDENCE IN NORMANDY 5

/\ / /,\ / i,\ -

,\ ~ =I l \ - -

\ \ . t , , - . \ - , - - - ~ \

1 0 3tr s ; . ~ _ q l FIG 5

Creully. Site plan showing on the W side of the bailey the 12th-century domestic buildings and lateraccretions. The existing ramparts an: largely post-medieval. SUn 9 M Wld drllW E. A. I 993


8/29

E DW AR D IM PE Y

FIG. 6Creully. Th e chateau from the SE.A 2th century single-aisled hallwith chamber block adjoining,much altered in the Middle Agesand since bu t essentially complete.Photo Linda Grant Courtauld nstitute

arcade - appearing end on in this elevation - largely survives, and retains onecomplete clerestory window towards its N. end . Below and to the right the masonryincorporates the end-wall of the aisle; although the pitch of the roof is not clear itsupper limit is marked by a string course.Ins ide, the capitals and springing of three arcade piers have been recentlyexposed within the thickness of the blocking-wall Fig. 7 , in addition to a series ofground floor windows and the main doorway from the bailey.

The second building consists today of a vaulted basement with interiordimensions m by 6 m and projects for 14.6 m beyond the N. end of the hall, itslast bay being flanked on the courtyard side by a small vaulted protrusion Fig. 8 .Until c. 1950 the five quadripartite vaulting bays to the S. and the remaining fourto the N. slightly different were divided by an original cross-wall. The interior islit by eight west-facing windows, and retains a semi circular recess in the E. wallwhich may have been an original fireplace. Since at least 1818 the undercroft hascarried like the hal l, an open air terrace but features such as the blind arcading


9/29

SE IGNEURI L RESIDEN E IN NORMANDY 53

FIG 7Creully. Pier and capital showing the springing of twoarchcs of the arcade between the main part and aisle ofthe hall. Notc thc variety in ornament Photo Linda GrantGourtauld nstitute

facing the terrace on the outer wall of the hall show that there was formerly anupper floor divided by a cross wall standing on the one below. A small vice in theintervening wall thickness may have linked it directly to the hall but the mainentrance was probably associated with the vaulted cellar at the other end the baseeither of a landing or a first floor porch


10/29

54 EDWAR D Ii\1PEYEqually important are the domestic buildings which survive in the bailey ofthe castle at Bricqucbec Manche , a possession of the Bertrand famiJy from theIOlh century until 1348.8 10 this case th e single-aisled hall, datable on the basis o fstyle an d sculptural detail to c. Ilgo, survives almost complete. As at Beaumont,

th e shell of the building ha s been preserved y the flooring-over and subdivision o fthe interior in the later Middle Ages; the arcade, exposed y the demolition of theaisle and blocked to Conn t he n ew north from Fig. I I ca n also be glimpsed in thefirst-floor bedrooms. The main vessel was lit by three windows in the S. wallmodified 1300 and a pair in each gable Fig. I2); whether there was a clerestoryover the aisle remains unclear. The cn d of the hall communicatcs with a twostorey construction of the same build, 6 m wide internally an d formerly projectinginto the bailey by at least the width of the aisle. The ground Aoor, linked to thc hallby a doorway close to the aisle Figs. I 2, 13 , an d lit by loop-windows in its W. wall,was probably a service-room. Above this was a chamber, decorated with blindarcading and adjoined by a latrine in the thickness of the S. wall. Th e chamber wasreached by an impressive doonvay overlooking the body of the hall Fig. 12),approached by a stone stair, as also in the near-contemporary English halls atOakham Castle Rutland an d Warnford Hants. . A spiral staircase in the thicknessof the wall led from the chamber either to a parapet walk or a gallery across thewidth of ha11- possibly both.

The second building lies to the NE . Fig. 10), positioned so that a doonvay atfirst-Aoor level could have opened into the aisle of the hall. Represented only by itshalf-sunken undercroft , its original fonn is a matter of conjecture: dearly adomestic building, its identification as a chamber-block relies on its position at theupper end of the hall and in the raising of the main Roor over a basement,conforming to the pattern outlined above. Although it has no easily datable dctail,it is necessarily earlier than the vault ofc. 350,which blocks the original windows.The last in the sequence survives in the commune of Bameville-Ia-Bertrannear HonAeur Figs. 13-15 .9 Here a chamber-block of standard typ dating fromc. 1200-1225, substantiaHy intact, is abutted at right-angles by a contemporaryhall, once probably linked via a survivingdoorway to a service block at the far end.The overall width of the hall is shown by wall-scars on the E. gable to have been nomore than 9.20 m, but a set of vertical, horizontal an d diagonal slots in thestonework of the W. gable respectively intended to house a wall post, a t iebeaman d a brace , show that it had two timber arcades Fig. 14, 8-B : given the lengthof the building [3 m) there were probably two bays. Consisting, therefore, of adouble-aisled hall an d a chamber-block ofstandard type, the complex is remarkablefor the use of the formula at a site of relatively minor status, its double-aisledconstruction and its timber arcades.

The only pre-122,5 hall surviving in isolation known to the present author isthe Echiquier at Caen, which, although half destroyed in 1944, has since beenI For historical notes see C. Barbe, L _ tk nmur I06Q 1350 11tbc: de Mailrlse, U n i \ ~ r s i l t de Caen,C.IQ80 . he building was finl inlcrpreled alon.g lhese lines by M. ~ l a r a Dupuis.


11/29

M I.... c l l llll I't ;a' I h c l100 CJ

p t ~ c l 6 0l 1last III c 1475 Pllast \ c 15JO

Pb;aw V c 164O

181h lOlh c< luI )' 0

Q

, - _ - - - - ~ ' : . . ~ - - ~ FloC ~ l I y . c,, nd oO pl.n of,h dom,,,i b i\din ,owin, p , o ~ , i o n . ph u , . S oM d . A I992


12/29

E DW ARD IM PE Y

Metres o 5.

IG 9Creully. Transverse section section line marked B-B, Fig. 8), facing S The roof-crease on the polygonal tower 15th century) marks the pitchof the 12th-century roofover the main body of the hall, originally open fromRoor to rafters the existing vaults being insertions ofc. 640). To the left was an aisle, openin out through thearcade, of which one capital and the springing ofan arch is shown. The vault to the far right, ormerly carryingthe seigneurial chambers in the position of the existing terrace, is original. Survepd and draWfi . A I r993restored to what must approximate its 12thcentury form Fig. 16 TraditionaUyand plausibly attributed to Henry I, this is known from excavation to have beenbuilt onto a complex ofearlier residential buildings - induding one which may beinterpreted as a chamber-block. iO Measuring gO 7 m by 11.02 on the inside, it hasa richly ornamented original entrance at its S end and is lit by six single-lightwindows o n each side and another in the S gable. Although assumed by M. d eBOilard, in line with the first-floor hall model, to have had two floors in its originalform, 11 this is dearly not the case: not only is there no archival or structuralevidence of original ground-floor windows or of a doorway to the floor above, it isfundamentally similar to a series of 12th-century unaisled halls in England10 See M de Boilard, eM tau Cam Caen,1979) 63, and E Impey, op. cit. in nOle 1,84-85de Boilard, op.cit. in note 10,70-72.


13/29

S E I G N E U R I A L R E S I D E N C E IN N O R M A N D Y 57

o 50Metres fIC.IO

Bricquebee. Site plan, showing the 14th-eenlury dfl ljun t o rampans mixed periods), the ha n B an d thechamber-block A . Adapled from Du Moneel, Th., Album u t A ~ l U du Diparlnnnlt dt t lbtouValognes ,843,fig.1.notably at Monks Horton, Dover Priory, Minster Court Kem), Sherborne OldCastle, and Old Sarum Wihs.).12 Thc romanesque timbcr arcades re-uscd in thc13th-century manor-house at Rumesnil Calvados)13 vcry probably represem anaislcd hall of the pre-1225 period, bu t only excavation could detcrmine its site an dform.As in England, chamber-blocks surviving in isolation are more numerous,owing to the comparative ease with which they could later be adapted. The

It fo r further detail on this point IIeC E. Imper, op. cil. in n ole I, no le 17. This view ab o held Jean I\laqui,CUka FAuittUsIk14 FrRMi Midiir 2 Vols Paris 199 ,1993), 81-82 .s E. I m ~ e Manoir de Rumesnil, raussernent appdc a r a n ~ all Dimes ; un Hall mCdiC\-al;l bas-e6tt aueentre de Normandie , lL Sot_ A tiq J{- LXI IggS), I 19 - 144.


14/29

EDWARD IMPEY

12th cenlul) _c.l300 14th nlUI) _Posl-medieval Metre ;

N

t lG

Bricquebcc. Plan of the hall at g r o u n d ~ f l o o r level and N. elevation. SUrt 9'ed and drawn E. A. I., 1992 1994.earliest, recently revealed by excavation 14 and tentatively identified below as achamber-block on the basis of its p la n, stood to the west of the keep at DomfrOnl Orne). Roughlycontemporary may be the fragmentary structure at t t e v i l l e ~ l A Nisscn:Jaubert, Domfront chateau): la chapell< Sainte-Catherine , Bilan itnt fi'lU(de la gion de Bam-}formondie, l im:lion Rigionok des Affaires Culturtlks ServU;e Rigionol de L'Ardliologil. Caen, [993),68 6g


15/29

SEICNEURIAL RESIDENCE IN NORMANDY 59

Metres I 0 5-IG .Bncqucbcc. Tran5\ocrK sectionoflhc hall ( lCCbon line marked A-A on Fig. I Facing W.). SMnx td ami mwtf..04./. 994.

Rue (Calvados). although its identification is not conclusive. The next in thesequence - in the Vexin Normand - survives at the priory of Jumieges atGenainville (Seine et Oise) ofc. 1140,15 followed by the remarkable building (laterused as a synagogue) ofwhich the undercroft survives below the courtyard of thePalais de Justice in Rouen;16 a nearidentical and contemporary structure wasdiscovered in the 1930S near the former Cour des Comptes, likewise built on anopen site within the city. The virtuany complete struCture at Loisail (Orne), in astraightforward manorial context. dating from c, 1180, is fonowed by the moreambitious example at Fontaine Henry, Calvados (c. 1200).18 The most importantof the remainder. all dating from the first two decades of the 13th century, can he

I> E. Impey, Tht 0riginJ and t w ~ t of N o ~ - C m W f l b u l 1 M o ~ a s i K Dtpmdnrnu it g/aJld and N1mIUltu J 1000- 350unpubliJhed docloral thesis, Univenity or(h,ford (199O), II, 50-62.11 D Pine, and P. Caillew:, 'I.e Mtimem hchr.llqUC de la cour du Palau de justice a Rouen',.o4TtIWL MMikaUXXll (Chronique), 4 ~ 3 - ~ 4 ; 'Architeclure civile en pierre a Rouen du Xle au Xllle ~ I e .o4rdIioi. MMiiwlt XXIV

(1n1), ~ 5 1 - 9 9 esp. ~ 8 8 ., C. Lanfry, La dtcouver1c d'une aIlebasse aRouen, ~ de l'ancienne Courdes Complcs', lLDe f4 e-m..o4.,iqItiliJ SeW Ijirinut (19)9),. 15s-s6; for a disc\Wion of manoria -Iypc complexes in an urban COllleX( seeR. Harris, TM Origilu fIII4 f j &,/UJI MtJiaIa T_ '-su O( 8lilw. SIoT .Jf unpublished dol::loraJlhois, UnivenityofOxford{I994}, l o - 4 ~II Imper, op. cit. DOle 5, llo-gl and 73-79.


16/29

6 E D W A RD I M PE Y

Barneville-la-Bertran. View of t he h ou se from t he SE. Th e early 13th-century chamber-block stands t o t heright; the multi-period structure to th e left incorporates thc remains of a double-aisled hall with timber arcades.Photo A. 1., /994.

listed as follows: Ardevon Manche),19 D ou vr es l a Delivrande Calvados),20 Glossur-Risle Eure), Martin-Eglise Seine-Maritime) and Honguemare-Guenouville Seine-Maritime; perhaps c. 1240).21 To these these m ig ht b e added some possibleexamples from Guernsey and Jersey, including th e romanesque Great Hall andthe later Chapel of St George at Mont Orgeuil and th e chapel at Samares Manor Jersey).22

The importance of t he h al l and chamber-block tradition in Normandy in th epre- 1225 period is confirmed by its persistence and continued development, in theface of an increasingly dominant alternative, throughout th e following century. Adirect derivative survives, for example, at Rumesnil Calvados),23 while th e vastopen hall at Le Neubourg Eure)24 must have been part of a comparable complex;standing remains of a double-aisled ground-floor hall of c. 1300 Fig. 18) andchamber-block of c I 220 Fig. 19) both survive at th e episcopal manor at Douvresla-Delivrande Calvados).25 Still later structures, clearly derived from the samepattern, have also been identified. It is also wort h not ing in this context that th e

19 E. I mp ey , L e p ri eu re d Ar de vo n: p r ~ u historique et architectural , Bull. Soc Amis Mont Saint Michel X CV I 1991),24-51.20 E. Imp ey , o p. ci t. i n note 5, 58-66.21 Recorded b y t he author 1993 94.22 For plans an d descriptions seeJ. McCormack, hannelIsland hurches Chichester, 1987),figs 99 an d 107. E. I mp ey , op. cit. i n note 1324 Main dimensions an d S elevation recorded by the author, 1994.25 Clarified by fieldwork in 1994. Th e hall measured 14.35 m b y 19.45 m, a nd h ad two arcades of four bays.26 M.J.-C. Bans, pers. eomm.


17/29

SEIGNEURIAL RESIDENCE IN NORMANDY 6.

FIG. 14

Barneville-la-Bcrtran. Transverse sc:


18/29

E D W A R D I M P E Y

F I G . 16Th e Echiquier at Caen. A ground-floor hall dating f rom r eign of Hemy

Photo E A 994

ANALYSIS O F T NORMAN EVIDENCEPresent evidence would no t in itself allow the evolution of th e ha ll and chamberblock ensemble as f ou nd i n Normandy to b e w or ke d out in any detail, but certaindevelopments parallel to those i n E ng la nd can be observed. Most obvious ar echanges in th e relative siting of th e hall and th e chamber-block. At Caen th eearliest of th e known sample the hall haphazardly associated wit h t he c om pl exof earlier buildings, i nc lu di ng t he possible chamber-block. At Beaumont-IeRichard th e buildings ar e on approximately the same alignment an arran gem en tancestral to the linking of th e two buildings longitudinally, as at Stamford Castle Lincs.),27 but whi ch never becam e widespread. The arrangement at Creully stillmore idiosyncratic not least as, although contiguous th e main link between th ehall and chamber-block was still external. Nevertheless, both arrangements displayan appreciation that for aesthetic and practical r ea so ns h al l and chamber-blockshould form a coherent structural unit.

The complexes of c. 1190 an d 1220 at Bricquebec and Barneville-Ia-Bertran,however are m ore directly ancestral to the arrangement whi ch became increasingly common in t he 1 3t h c en tu ry at least in England whereby th e chamber-blockwas placed at right-angles across th e upper end of th e hall, frequently m irro red b y

27 C. Mahany Excavations at Stamford Castle, 1971-6 Chiiteau Gaillard 8, 223-45.


19/29

SEIGNEURIAL RESIDENCE IN NORMANDY

r ; ~ l:J oDco_J_Ll :== ._. _ JJ

3

w ~ ~ ~ ~ : ~ ~ t ~

~ r = : = = ~5

4

--=1. I~ . 1 . . - - _ . _ ~------~ : r T ~i:: eS_ 10t::-.- ....Tee ..

11.. T .::::; T -10 0 30Metres .,. ., ;OoO

fIG . 1 7Selected ground-Ooor halls i l England and Nonnandy. Simplified plans, shown at the same scale.

KEY: Medieval fabric of/extant at/date given: solid black standing or exca,... ted; broken line inferred. Laterfabric: stippled(,) Beaumont-Ie-Richard (Calvados), e. , 50 . E. A. I., (7) Hereford, c. I go . ReoonsU UCtion. Afier1993, E.A.I., m(I) Blair, ,gB7(2) CreuUy(Caivados), c. 1160. E. A.I., 1992, E. A. I., (8) Oakham Castle (Rutland), c. llgo. After

993 VCH RlIl II J(3) Caen (CaI,oados),c. 1125.;\ rebuilt after 9++ After (9) MinsterinThanet (Kent), c. 1150. Adaptedfro>de-...ux, 972 from Kipps, 1929(4) Bricquebc:c (Manchc:), e. , ,go. E. A. I., 991, 99](1) (10) Wamfor-d (Hants.), c. 200. AfterVCH u t(5) Douvres-Ia-Dclivrandc (Caivados), c. 300. E. A. I., (11) \Vinehester, cutIe (Hants.), 1222-35. Aftert99](2}, and funhc:r ....,ork E. A. I., 994 VCI- aIfb(6) Rumanil (Calvados), c. 260. Rcc:onsU UCtion. (12) Winchester, Woh-eser Palace, east haU asE. A. I., 199](I}, E. A. I., t gg8 rebuilt I 35-8. After Biddle, ,g86


20/29

EDWARD MP[Y

8

4c ~ f-.[

7

3.t= u l

6

2

n L

FIC.18(ahootaJIdJllCiIItl

Selected chamber-blocks in England and Normandy. Simplified plans, with some section$ and elevations,$hown to the same scale {same as Fig. 17}.KEY: A denotes ground Roor or undercroft: B denotes lSI floor. All sectioned Willis CIC. black. Otherwise, asfor ''8:.17.

I) oothbyPagndl (Lines.), c. I 180. E. A. I. and Cogges (Oxfordshire), 11,)0-80. MtCT Blair, .g82.Harris, fonhcorning. (6) Gl O ITlOnl (MonmouthshiTc), I ~ O I - .(2) Bunon Agnes orb.), c. 11]0. After Wood, 1963. Knight, IgSo.(3) Cambridge, Menon Hall, t I ~ After R H ~ I (7) Jacobstow(Cornwall), 1180-1200. After BerWord,1974-(4) Christchurch (Dorsel), c. 1160. Afler Wood, 1964. (8) Eynsford (Ken ) . After Rigold, 1971.a second chamber raised over the sClVices at the opposite end. At Bricquebcc,although the siting of the buildings was partly dictated by the line of the curtainwall, they are closer to the right-angled arrangement than the axial; crucially, theentrance to the chamber was in its sidewaU and nOt, as at Beaumont-leRichard,in its gable end. Equally significant is the placing of another chamber, smaller butclearly important, over the services; as this struCture projected into the courtyardand had a roof at right angles to that of the hall, it forms one of the earliest trueknown cross-wings. At Barneville-la-Bertran, the 3thcentury chamberblock isplaced across the upper end of the hall in whal was to become the standardmanner, although, curiously, it was linked to the body of the hall only by anexternal stair. vVhether there was a second chamber over the services at the otherend of the hall is unclear.When we come to consider developments in the design of each element, thereare, again, similarities to those observed in England. At the Echiquier, the mainexternal doorway is in the gable, an arrangement which it shares, for example,


21/29

SEIGNEUR T L RESIDENCE NORM NDY

rr '--:::, l I ~ . 1 , ~ .

[:;J lll12

16

. ~ ,

10tEl. 1.1A X _13

15

17

~~ ::

. l,14 ,

18 19

C:JJ , leli 2I 0 300MHres . .. .. .. .. . I209 Stamford (Lincolnshin:), as in late 12th century(showing hall also). After Mahany, [977.10 Strood (Kent), c. [220. Aftcr Rigold, 1962.(I I Ardevon (Manche), c. 1220. E. A. I., 199L(12) Beaumonl-Ie-Richard (Cal\'arlos), c. 1140-50.E. A. I., 1993.(13) Bricquebec (Manche), c. 1190. E. A. 1., 1993(2), 1997.(14) Creully (Cal\'ados), c. 1160. E. A. I., [995.15 Domfront (Orne), c. I110. After Nisscn:Jaubert,

1994

Douvres-Ia-Delivrande (Calvados), c. 1220.E I., 1993(2).''-ontaine Henry (Calvados), c. 1200. E. A. I.,[993(2).Clos-sur-Ris1c (Eure), c. [220. SurveyedE. A. I., 1993.Loisail (Orne), C 1180. E. A. I., 1993(2).J\Iartin-Eglise S c i n e - ~ b r i t i m e . SurveyedE. A. I., 1994.Rouen, Lasynagogue'. After Plue, 1994.


22/29

66 EDWARD IMPEVwith the th-century hall at Weslminslcr.28 At BcaumomleRichard the positionof the main external doorway is unknown, but at CreuUy it was placed at theextreme S. end of the aisle. A necessary prerequisite fOT the evolution of the CTOSS-passage, this suggests that it may have had a plan of the highly evolved type ofwhich Oakham Castle (Hg. (7) is the earliest ncarcomplcte example in England,whereby the main door was up against a gable wall pierced by service doo ....vays.At Bricquebcc and Barneville, although the main external entrance has been lostalong with the aisle, a service door in each case survives. hal can be observed of the development of the chamber-blocks clearly alsoconforms to the English pattern. The stone structures at Domfront and atVatteville-laRue of c. I 120, if correctly identified, represent the classic chamberblock' in its simplest two-storey form,29 while the building at Beaumont-Ie-Richardrepresents the same form in its maturi ty, not only with the habitable part raisedover a storage room - in this case barrel-vaulted - but divided into two unequalparts, exactly as found in the archetypal English buildings at Boothby Pagnell(Lines.), Christchurch (Dorset)30 and Strood Temple (Kent) (Fig. 18).31 It is worthnoting that , although merely a decorative feature, bl ind arcading at the latter isemployed in a strikingly similar way to that at BeaumontleRichard. The chamberbuilding at Creully, where only the lower part survives, seems to have conformedto a similar model, complete with a cross-wall at both levels. A refinement here,however, the extra vaulted cell that may have carried a first-floor porch - alsoprobably the explanation for a near-identical feature at Fontaine Henry, datingfrom c. 1200. In common with the 'synagogue' in Rauen, the t\\'o floors at FontaineHenry were linked by a spiral stair in the thickness of the wall: similar arrangementsexist in England at BuTton Agnes (Yorks.)]2 and Christchurch Castle. The laterexamples, including the late 12thcentury 'Auditoire' at Loisail, are, as in England,rather simpler, in general having only one first-floor room, perhaps in response toa routine inclusion of a secondary chamber over the services.In general terms, some features of the Nomlan buildings suggest that Normanpractice, at least in structural terms, was in advance of that in England: inparticular it is worth noting that the stone-arcaded halls at Beaumont-Ie-Richardand Creully date from a generat ion before the nearest equivalents in Englandsomething that might be attributed to the tradition of stone arcading inecclesiastical architecture evident since c. 1000 and which by at least 1200 hadextended to barns.]] In addition, the building at Domfront, if indeed it was achamber-block, is the earliest example in the AngloNorman world to display whatwas later to become such a typical ground plan.l H. 11 1 Colvin, HisltJry o{/k I i i ~ K ll ...,.b: t Middh gn (London, 1963), I, 46. For Domfront :lee A. Niucn:Jllulxrt, 'Lo: Ch l.ICllu de Domfront au ,\-Ioren Ago:: approche archrologique ethistorique', 147-61 in Fajal (cd.). M lIftT CMtmll miJiimJ Sodttt historique el archi:ologique o l Orne,

~ I i : m o i r c s et Documents No. I (Alelli;Qn, 19 )8); for Vatte...ilk-Ia-Rue,:lee A-M. f1amba.-d Heridw:r, 'La CuisilW:du Iogis : l e i ~ } c U I i a du V)eux Chilo::;lU do: Vanevilk-Ia-Rue (Sdnc-J,laritime)', 131-46 in rajal (cd.). op. cit. M. E. \\00d, 17tt fAgliJJI MtdiMwll Hf1IU (London, 19l1.f), 17-19. S. Rigold, T...pIIM_ S t r o N { H ~ I S O IgOO)., . M. E. Wood, ~ g w OliMa_Haut HMSO 1956 . n.c earliest example being perhaps the magnificenl ~ C n - b a y c d construction in the: grounds ofSaini-Wandrillc(Seine.)'Iaritimc).


23/29

SEIGNEURIAL RESIDENCE IN NORMANDYAN ANGLO-NORMAN TRADITION?

That the hall and chamber-block tradition was common to b ot h E ng la nd andNormandy in the century or so after 125 raises the question of wh et he r it was inany sense peculiar to these regions. We know that apartments differentiated in theMiddle Ages and identified t oda y as halls and chambers were common toseigneurial residences over most of western Europe, but was t here in fact anydifference in the Anglo-Norman structural response to these requirements, or didit co nfo rm perhaps with slight regional variation to a much more widespreadtradition?

If we look at seigneurial buildings outside Normandy, we find, of course,much common ground. The grandes salles known or extant) at the c ha tea ux ofAngers Maine-et-Loire) 11th century) and Blois Loir-et-Cher) 13th century), atthe Palais de la Cite in Paris c. 1300) or at Montargis Loiret) c. 1300) havegenerally been accepted as close equivalents34 to those at Caen or Westminster. Toa large extent this is correct: we know that not only were they used for the exerciseof justice, but also that they were the setting for formal meals staged mu ch as inEngland, and that certainly in France the practice persisted throughout the MiddleAges. But if we look at the p lan of these rooms in more detail, there are tellingdifferences notably in the absence not only of the Anglo-Norman layout of theservice and access routes but of a consistent alternative. More importantly, thelater examples cited represent a tradition in the organization of seigneurial houses,observable and probably dominant in France and Germany from the 12th centuryonwards, in which there was no single communal hall, but separate halls providedfor the upper and lower ho useh ol ds s up eri mp os ed within the s ame s tru ct ur e.Important early examples include the bishops palace at Paris, put up by Mauriceof Sully afte r 1160 and known from pre-Revo luti onary records,35 while aRomanesque structure of t he sa me form is also implied by the two-level chapel atLaon Aisne). A variation of t he same s to rey ed -h ou se pattern is found inGermany as for example in the 12th-century domestic buildings at Milnzenberg Hesse)36and the Wartburg Saxony).37

The common use of the storeyed-house formula on the Continent outsideNormandy does not, of course, rule out the parallel use of practices much morelike those of England and the Duch y. B ut a lt ho ug h it is da nge ro us to argue th at acertain form of structure, as with any artefact or practice, did not exist in a certainarea in a certain period, the Continental use of anything akin to the hall andchamber-block formula outsidc Normandy is it seems, extremely rare: moreover,those domestic buildings in neighbouring areas which do more than superficiallyrecall the Anglo-Norman pattern can be attributcd to Angevin builders to whom it \Vith reference to the last three, see Colvin, op. cil. in note 28, 44 fig. 9).s E. Viollct-le-Due, Ditlionnairt rauonni dt l arckittcturt ftanfaist du t au XVlt siklt Paris, 1858-68), Vlll, l.l,-16and 70-716 B.JOS1 The castle at MlInzenberg Hesse), and its relationship to later twelfth-century eastle, , in G. :\oIcirionJones et al. eds), lt StignliJrial tsidtllct in Euro M forthcoming). K. M. Swoboda, Riimiu:M und RomanUtM Palaslt: int arckildflurgesckitktlickt UnltrSutkung Vienna, 3rd ed., 969); Asche, t Wartbmg: GtJCkitklt und Gtslall Berlin, 962).


24/29

68 E D W ARD IM PE Ywas familiar. For example, the superb double-aisled grenier at Angers38 by thelate Middle Ages in usc as a store-building bu t surely, in origin, domestic datesfrom the generation after I 54 an d ca n be attributed the patronage of Henry11.3 9 The Sieur dcJoinvillc s description or the building at Saumur (Maine et Loire)used by 5t Louis in 124- suggests that it was an aisled hall, bu t altributes it to thegreat Henry II ; the grande salle of the palace of I Ombriere at Bordeaux(Gironde), known from dClailed 18th-century plans40 to have consisted of a navean d single aisle with a four-bay arcade between, is, once again, almost certainlyalso ofAngcvin origin. The same can be said of the palace of the counts of Maineat Le Mans (Sarthe), known from a plan of c. 175041 (which contained a giganticdouble-aisled hall ofseven bays), reminiscent in its close-set posts ofarchaic Englishhalls such as Henry I s construction at Cheddar (Somerset),42 an d also or laterexamples such as Leicester Castle,43 an d Cheddar as rebuilt by KingJohn. 4 4 TheAnglo-Norman plan of the ramous hall at Poitiers (Vienne), complete with pairedservice doors opposite the dais end, built at the en d of the 12th century by Eleanoror Aquitaine,45 ca n be similarly explained. Thus in the period c. 1125-1225,present evidence suggests that the strictly defined hall an d chamber-block formulawas particular to the Anglo-Norman builder.

THE ORIGINS O F T HE ANGLO-NORMAN TRADITIONIr a specifically Anglo-Norman tradition did exist in the 12th century, anexplanation is required. Among the possibilities are the following:

that the tradit ion represents an ancient pan-European custom extincteverywhere by the 12th century except in Normandy;2. that the tradition evolved independently in England an d Normandy beforethe Conquest; that the tradition emerged in Normandy an d was transrcrred to Englandafter the Conquest;4. that the tradition had its origins in England and was transrerred toNormandy arter the Conquest.

The first hypothesis, although plausible, is no t supported by an y specificevidence, while the second is inherently difficult to test. With regard to the third,although excavation in Normandy has revealed pre-Conquest structures belonging

:J A. Mussat, L H6pital Saint:Jean a Angcrs , Q n ~ i s Archiof. Franet, eXXll (1964),79 an d 85. His/oire dt Saini u ~ i s ed. N. M. de Wailly (ParIS, 1878), 54. On thc inlcrpretation of lhis passage see M. W.Thompson, 71a RistO/1M Costu (Cambridge, 199 I , 44-46. Archives eommunalcs dc Bordeau , XU/8;]. Gardelles, eM/taux du ml J(II agt dans Franet sud-ouesl (Geneva,

1 ~ 7 ~ , 05-07 an d fig. 26, pI ix . Archives Depanmentales de la Sarthe, cat.m). 187; R. Triger, Th6tcl de ville du Mans (1471-1896 lJlltHisloriqut tl archiowgUJIie du MaiM 43, 1898, I13-88, 133; A. Renoux, Palau midulI


25/29

SEIGNEURIAL RESIDENCE IN NORMANDY 69to the Gennano-Nordic vernacular r r a d i t i o n , ~ and complexes of the 'hall andchamber type dating from the Merovingian, Carolingian or early Ducal periodmay await discovery, nothing (with the possible exception of the curious timberbuildings excavated at Grimbosq (Calvados) 4 yet heralds the Anglo-lormanpattern as defined above. Nor can this be said of thc more ambitious complexes,such as those reprcsented by the enigmatic 'Palais de Guil laume' at Lillebonne(Seine-Maritime), the fragments excavated in the castle at Caen, or dwellingscontained in or including towers, such as Duke Richard I's (943-97) at Bayeuxand Rouen, or the partially surviving examples at Ivry-Ia-Bataille (Eure) (c. 1000)and Avranches (Manche) (c. IOjO?).48 Nevertheless, bearing in mind that its firstidentifiable appearance in Normandy dates only from the I 120S, the pattern couldstill be assumed to have originated in Normandy after the Conquest , were it notevident in England well before 1066. The evidence for this is to be found in bothliterary and archaeological sources, recently summarized and analysed by Blair.49The functional prerequisite - the distinction between hall and chamber - isimplied as early as the 9th century, notably in King Alfred's own works and in hisbiography by Asser, in which, for example, he writes de aulis et cambris regalibus','marvellously constructed of stone and WOOd .5O Structural evidence, revealed byexcavation at fI.'Iucking (Essex) >l and Cowdery's Down (Hants.), >2 suggests that thedistinction was being made as early as the mid-Saxon period; houses at both siteshave subdivisions, at one end of hall-like structures, which can be interpreted aschambers. >3 Examples dating from after c. 1000, however, display characteristicsmore blatantly ancestral to the Anglo-Norman pattcrn in its mature form, in whichhall and chamber are housed in dist inct structures - a development that in t imberbuilding was probably regarded as a refinement. Examples include Sulgrave(Northants.), >4 but the clearest example is probably that of Goltho (Lines.), J. cMaho, La MIIlU de Miroilk ( X ~ X l l t siirw) (Rouen, 914), esp. figs 7 and 24; A. Renoux,lWlImfJ DII PIIf4is DIIuJa I\J.isde Diat{Paris, 99 ). ,.U Dccaens, La Moue d'OIi\ et lGrimbosq (Cal\ ados): residence scigocuriale du }G c sittlc , ArdWL MWimluXI (198 ), 1 6 7 ~ 2 Q , esp. fig. I. The excavator, however, interpre15 the larger building as a t w o - s t o r e ) ~ d Structure. For the lower at Bayeux $CC A. RenOllx, 'ChAteaux et rtsidcoccs f o n i ~ s des Dues de Nonnandic aux X et XIsittlel in AClcsdu I ,erne Congrt5lntemational d'ArchColO$ e mtdi< :vale tenu i Caen, October 987, pp. 3-24,p.12 ; Rouen is mentioned by Roben of Torigny and attnbuted yhim to Richard J (The O t r l l 1 l ~ U RIJbnt Torigny ed. R. H o ~ e 1 t , C J l T l 1 l ~ l u ri kReigru St1JMrr, Hmry ftuld R ~ I u I T l I (Rolls Series, LXXXll , ,886), v, 106);for J ~ - l a . B a t a i l l e see E. Impey, jh e Turrisfanresa at Ivry.laBalaille, Normandy , in G. Mcirion:Jones et al (eds),The snpteurW Ruidnu:t ill Europe (forthcoming); for Avranchcs, see E. cHericher, Avran,hin rnD1lunrmla d ltisl


26/29

7 E W R TMPEYexcavated in the 1975. Here the Period V dwelling extant c. [000-80 consistedof a timber-built single-aisled hall measuring approximately 16 m by mexternally, accompanied by a separate building just over a metre from its E. endwith dimensions of 7.5 m by 9 m, interpreted by the excavator an d others as abower or chamber. 55 At Deddington Oxon.), a stone chamber-block and timberhall seem to have existed side by side by the 1 0 5 0 s. 5 6An implication, supported once again an examination of other Continentalevidence, is thus that the hall an d chamber-block paltern ha d its origins in aspecifically Anglo-Saxon practice,57 taking o n a n An gl o-No rma n di men sio nthrough its transportation across the channel after 1066. This raises the question ofwhy, when in the fields of ecclesiastical an d military architecture Norman practicesuperseded the English, did the Normans adopt an alien pattern whe n it came totheir houses? The answer ma y lie in the relative degree to which coherent traditionsin these various fields ha d b ee n developed before the C on qu es t: by 1066, AngloSaxon church buildings ha d rarely achieved great distinction or great size, bu t inNormandy, thanks to links with a long-established Continental tradition, they ha ddone both, an d the Normans ca n have felt that they ha d little to l ea rn from t heirnew subjects. the field of domestic architecture, however, the Duke and hisfollowers were confronted with a stock of palatial complexes perhaps every bit asimpressive, in their way, as anything which they ca n be shown to have possessed inNormandy, among which were those at Winchester, \estminster an d Kingsholm Glos.). The Winchester royal hall proved inadequate during the reign of Henry Ibu t that at Kingsholm,just lO the north ofClouccstcr, survived well into the reignof Henry Ill, while it has been argued that William Rufus s prodigious hall a tvVestminster replaced an Anglo-Saxon timber structure of comparable dimensions The same impression could as easily have been encountered by those lowerdown the social scale. As the Norman seigneury seem to have ha d no coherenttradition of t heir o wn in purely residential building, the highly evolved AngloSaxon pattern may have ha d an immediate appeal.ALTERNATIVES TO THE HALL AND CHAMBER-BLOCK PATTERN INNORMANDY, 1125-1225

The hall an d chamber-block pattern was not, ofcourse, the only or necessarilythe dominant arrangement e mpl oye d b y Norman seigneurial builders in an yperiod. Th e form of the Phase building excavated at Mirville Calvados) aloneshows, no t surprisingly, that the use of Germano-Nordic vernacular buildingtechniques persisted at least beyond 1100 even if the planning could haveconformed to the Anglo-Norman pattern); in addition, the Continental storeyedG. I k r t ~ f o r d CoI/lvx The Del. elopmtn/ r llIl Eor{J MdinJal MaMr 850 1 5 London, [987), tSp. fig. 75 and

Pg7[-84.Th e late Martyn JoJ:N pets. comm.; see also R.J. Ivtns, Deddington castle and Od o of Bayeux , OXQniensiaXLlx [984 ,IO[-[9S.Jamts et aI., An early mtdieval building tradition , Archatol.]. CXLl [984), [811-11 [5Colvin, op. cit. in note \18 4\1-47.


27/29

S E I G N E U R I A L R E S I D E N C E IN N O R M A N D Y 7house , in its purely domestic form, was already being introduced into Normandyfrom France proper by c 12 I 59 To these, the AngloNorman formula probablyrepresented a straightforward alternative, bu t its relationship to its most prominentrival the defensible residential lOwer, i.e. buildings usually referred to as donjons ismore intcresting. If, as had almost certainly been the case at Ivry-la-Bataille an dRauen, the larger 2 t h ~ c e n t u r y towers such as those at Falaise Calvados), Cacnan d AleniYon Orne) at least had the capacity to serve as en ti re residences, theycould c1carly also offer a straightforward alternative to the arrangement adopted atBeaumont-ie-Richard, Bricquebec an d Creully, or implied at Fontaine Henry. Butthe fact that few residential towers contained the cntire residence lcd, in practice,to some bl urrin g of the distinction: mid-12th-century Norman tower residencessuch as at Brionne Eure), which contained only one or two habitable rooms, mayeffectively have functioned as defensible chamber-blocks, an d been accompaniedby other residential buildings, even ifnot necessarily by a hall of Anglo-Normanform. Such hybrid arrangements may in some cases have remained in usethroughout the Middle Ages, but by the end of the 12th c en tu ry , a t sites where allthe main accommodation was within a tower, their residential role ma y frequentlyhave b ee n s up pl an te d by a mo re practi cal hall and chamber-block complex in thebailey. Certainly this is suggested by the English experience, an d to some e xt en t byarchaeological evidcnce at Domfront and Falaise,60 an d ma y have o ccurred atCaen, where Hcnry I buill b ot h the Echiq uier an d a gigantic keep, as early as the1120S

CONTINENT L INFLUENCES ON ENGL NDThe possibility that the 12th-century Norman seigneury ma y have adopted an

English model for the building of their houses does not, of course, mean t ha t p re Conquest C on ti ne nt al practices were not also transferred to England: bothContinental planning an d C on ti ne nt al b uil din g t echn iq ues were to have a majorimpact on English domestic architecture after 1066.The most obvious building type to be introduced was the keep or doryon itsfirst manifestation in its mature form b ei ng e ith er the Whi le Tower or Colchcster Essex). But the donjon was also accompanied by other variations of the Continentalstoreyed house, now identifiable as such an d not as prototype keeps, thanks to therealization that the mature four-square donjon was not a p os t-C on qu est dcvelopmcnt, but ha d existed in Normandy since about [ 0 0 0 . Examples of such buildingsinclude the curious tower at Chepstow Gwent) and thc double-pile building on thc

motte at Castle Acre Norfolk).61 The same broad interpretation ca n b e appli ed toother unorthodox buildings, including, for cxamplc, lhe gigantic west hall built

Y E Imp


28/29

72 EDWARD IMPEYby Bishop William Giffard 117-29) at Wolvesey Palace, Winchester,62 andperhaps also Le Puiset s double hall at Durham, which have much more incommon with the French and German palaces mentioned above than anything inthe hall and chamber-block tradition.

The parallel use of the alien storeyed house and the hall and chamber-blockformula in England produced results and compromises similar to those observableor implied in Normandy. The use as such of the residential tower incorporating allessential accommodation was particularly shon-lived: by or d ur in g the early 13thcentury almost every major ex amp le can be shown to have been acc omp ani ed bya hall and c ha mb er-b lo ck c omp lex in the bailey. Wit h the n ot ab le exc ept ion ofHenry II s deliberately impressive but already archaic keeps at Dover andNewcastle, related post- 150 structures can rarely be interpreted as entirehouses - a phenomenon anticipated at the episcopal palaces such as Wolveseyand Sherborne Dorset),64 where the keeps are effectively turriform chamberblocks associated with a vast residential complex.

With regard to structural innovation, the most obvious impactof the Conquestwas the increasingly widespread and competent use of masonry, not just toreproduce Continental forms, but to give new stature and permanence tocomplexes of the adopted hall and chamber type.CONCLUSION

In conclusion, it appears that mainstream seigneurial domestic architecturein Normandy in the perio d 125- 225 was essentially similar to that in England:the typical ensemble consisted of a two-storey residential block associated with adetached communal hall, usually open from floor to rafters and at ground-floorlevel, e ac h c le men t developing a rema rk ab ly consistent design. The apparentabsence of this specific response to the needs of a seigneurial household elsewhereon the Continent may justify the label Anglo-Norman . The explanation for thispeculiarity may lie in its basis in an insular tradition evolved in Anglo-SaxonEngland, transported to Normandy only after 1066. At the same time a n earliertradition of building storeyed houses persisted - most obviously in the form of thedonjon - and was transferred, in more than one guise, to England. KNOWLEDGEMENTSThe author would like to thank some of the many people who have contributed to theresearch and fieldwork on which this article is based. These include Mme DugucperouxArdevon), and Mme Dupuis Barneville-Ia-Bertran), Jean-Philippe Schnell Beaumont-Ie-Richard), Lady Nethersthorpe Boothby Pagnell), and Mme Hardy6 M. Biddle, lVoll> l9 t I l a H ~ I S O , 990).6 The plan ofGi fard s building at Wolvesey the west hall ) has an obvious similarity with that of the Wartburgand other German palaces where main rooms at two or more levels are fronted by an open corridor, and suggestsan interesting interpretation for this unusual building. e P uiset s north range at Durham could perhaps also beinterpreted in this way. See P \Vhite, SMrwrm Old as It HMSO, 971).


29/29

SEIGNEURIAL RESIDENCE IN NORMANDY Bricquebec), M. and Mme Riffi and M. Clairon Creully), M. Hirard DoU\ fes-laDeIivrande), Ie Comte et la Comtesse D Oilliamson Fontaine Henry), Mme Poot Glossur-Risle) M. and Mme Ripamonti Loisail), t> 1 and Mme Lcbeurier Rumesnil). Manythanks are due also to Mathew Impcy and Jean-Paul Corhasson for help with fieldwork denoted by initials M. A. andJ. P. C.), to Roger Ainslie for the loan of his theodolite,

and to Jeremy Ashbee, John Blair, General Pierre Garrigou Grandchamp Anna Keay,Linda Grant Jane Inskipp, Yves and Elisabeth Lescroart,Jennie Lowe and Gwyn MeirionJones for help and advice. Final thanks must go to the British Academy for funding most ofthe research and fieldwork through a post-doctoral fellowship 1992-95)and the DirectionRcgionale des Affaires Cultureltes for their contributions to fieldwork expenses.