Having analysed the essential characteristics of the archi• Is There a Pre-Romanesque tecture of ćhe Christian West between the 8th and ćhe llth c ent«ry, the a u thor p t>ts foru'ard hi s c onviction t hat a Pre-Ro»>anesque expression is preceding the Romanesque Style in Architecture'? style. The distingt>ishing features of the Pre-Romanesque architecture are biwxiality, the absence of correspondence between the interior organization of space and that of the Dr. Vladimir P . Goss exterior u>all-surfaces, the presence of hidden inćerior t
As we all k n ow , t h e n a mes of m e d ieval ar t s t y les the way in c h ich the architect communicates his con• are rather arbitrary. But, while we more or less under• cepts to his public. Definitions of the Romenesque em• stand what we mean when we use the word » R omanes• phasizing the points brought up i n t h e fe w p r eceding q ue« o r » Gothic«, there does no t s eem t o b e m u c h lines have been forwarded by scholars such as Baum, a greement as t o t h e m e aning anđ scope of t h e t e r m Beokwith, Brehier, Clapham, Conant, De Truchis, Dyg• »Pre-Romanesque.« Is there such a t h ing as a P r e-Ro• gve, Enlart, Focillion, Francastel, Frankl, Pevsner, Puig m anesque style in a r t , m o r e >particularly, in a r chitec• i Cadafalch, Saalman, Salet, and they have also found ture'? Is t h ere a P r e-Romanesque Ku nstwollen, some their way i n t o s uch p o pular h a nđbooks as Mc G raw• distinctive, »Pre-Romanesque«concept of the arcitec• Hill Dictionary of Architecture or H e len Gardner's Art tural space and mass which would j u stify th e use of Through th e A g es.' I f on e ad o p ts t h e p o s i tion j u s t the wo r d »P r e-Romanesque« a s a st y l i stic category stated as a starting point, one may, as a w o r k ing hy• covering all, or at least a considerable number of in di• pothesis, postulate >that the Pre-Romanesque architec• vidual arcihtectural creations in the period » before the ture will not ćisplay the logic and clarity we attributed R omanesque?«B r iefly, is t h ere anything i n t e r m s o f to the Romanesque as its essential characteristics. Let the architectural essentials common to all those groups us try to find out if a brief consideration of some among of »before- the-Romanesque-buildings:
33 V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITEČTURE7 PERISTIL 25/1982. (p. 33 — 50)
tLSSJEE C C IS'Đ>E
S JEVEE
tAESJKK A A
ZAtAD
TEOC ST
tAESJEE E E
I I '» S
O»>is,St. Peter, c. 1050— 1075, gro«nđ plan and sections 2. O>nis St. P eter, faeades and t r a nsverse section [author] [author]
F or a n u m ber o f y e ars i h a d b een i n t erested in a o f St. Peter at Q m i s (c . 1050 — 75). The building w as group of P r e-Romanesque monuments (109 on r ecord known to Jose Buig i Cadafalch, who consiđered it an so far) .in the Eastern Adriatic, traditionally known as offshoot of t h e » F i rst Romanesque Art.«' A superficial »Early Croatian.«' From t h i s g r o up, th e b u i l dings of examination of the exterior, characterized by the arched w hich date f ro m c . 800 t o c . 1 100, I s h ould l i k e t o corbel t ables, seems t o c o n t ir m P u i g i Ca d a f alch's select, as the first o b ject o f o u r a n alysis, the church c lassification (Figs. 1 — 3). But a n a n a lysis of t h e i n •
- "The bulk of the existlng literature on the Romanesque worth, 1963, pp. 11, 13; Hans E. Kubach, Ro»>a»esq«e Ar• m akes it impossible to collect all th e definitions of t h e chitecture, New York, 1975, pp. 11, 15; Nicolaus Pevsner, style. The broad definition proposed was arrived at through An Outline of E u r opean A r chitecture, Harmondsworth, the observatIion of the Pre-Romanesque and Romanesque 1963, pp. 56 — 57; Jose Puig i C adafalch, Le p remier art in the course of the author's own research, generously com• ro>nan, Paris, 1928, I, pp. 62~3 ; H o w a rd S aalman,Me• plemented by the ideas of the outstanding scholars in the dieval Architect«re, New York, 1962, p. 30. Also, The En• field. See specifically: Julius Baum, Ro manesque Archi• cyclopedia of World Art, 15 vols., New York, 1966, XII, p. tect«re inFrance, London, 1928, pp. 23— 24; John Beckwith, 319, aInd Helen Gardner, Art Through the Ages, New York, Early Medieval Art, London, 1964, p. 153; Louis Brćhier, 5th ed., 1970, p. 303. Le style roman, Paris, 1941, p. 40; Alfred W. Clapham, Ro• Although the investigation of a stylistic relationship bet• ma»esq«e Architecture in Western Europe, Oxford, 1936, p. w een the Romanesque and the Gothic Is not w i t hin t h e 23; Kenneth J. Conant, Carolingian and Romanesque Archi• scope of this paper, one should be made aware of the fact tect«re, H armondsworth 1959, pp. 11, 28— 29, 30, 42— 45, that many of the characteristics insisted upon as Roma• 57 — 59, 67; Ejnar Dyggve, History of Salonitan Christianity, n esque are found, or, in f a ct, fully blossom out i n t h e Gothic. See Frankl, Gothic Arc1iitecture,,p. 66 (»Even when Oslo, 1951, pp. 133, 136 — 137; Camille Enlart, Manuel d'arche• all the traces of th e Romanesque had disappeared, the logie francaise, 3 vols., Paris, 1902, I, p. 199; Henry Focillion, Gothic style was still a descendant of the Romanesque. It Art of the West in the Middle Ages, 2 vols., London, 1963, is a transformation of the historical style of totality into I, p. 62 ff.; Pierre Francastel, L'humanis>n ro»>an, Rodez, a style of .partiality.«) Also, Francastel, L'h«»>anis»> ro»>an, 1942, p. 104 ff.; Paul Frankl, Cothic Architect«re, Harmonds• p. 211 ff.
34 V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE2 PERIST1L 25/1982. ( v. 33 — 50) terior co mšined w i t h a n e f f o r t t o re l at e i t to t he exterior runs into a number of difficulties. The exterior of the aisleless building grows toward the squarish tur• ret with a pyramidal roof which acts as the centralizing element of t h e e n t ire architectural mass. The rectan• gular apse, just barely projecting, does not create any sense of longitudinal orientation. The five doubleheaded arches on th e l a t eral f acades emphasize the v er tical g rowth o f t h e m a s s d i r ecting one's attention t o t h e t urret. Th e o r ganization o f t h e i n t e r ior i s fla t l y i n contradiction with that of the exterior. The dome, hid• d en within t h e t u r r et , i s s o s m al l t h a t i t is ba r e l y noticeable. Its iimpact on one's experience of the space is for al l p r a ctical p u r poses megligible. The space is oriented longituđinally b y t he sem i c i rcular a p sidal niche a c companied by t wo small e r sem i c i rcular recessions. The individual mnšts of the interior do not corresponđ to the i n dividual u n its o f t h e e x t erior. The s quarish turret encases a dome, the r ectangular iprojection at the eastern end, a r ounded apse; the apsidiole niches are completely invisible from the outside. Moreover, the three-bay organization of the nave shows no correspon• dence to the five unit organization of the external wall surfaces. The interior and exterior seem to b e c learly separated in t h e m i n d o f t h e a r chitect. The r eaction of an a t t entive v i sitor u po n e n tering t h e c h u r c h • after an a n alysis of t h e e x t erior — is o n e of sur• prise.' To sum up , on e m a y sa y t h a t t h e b u i l d ing u n der analysis is characterized by the following: 1. Bi-axiality. Predominant r ole o f t h e v e r tical axis in the formation of th e m ass as opposed to t h at o f the l o n gituđinal axis i n t h e o r g anization o f space. 2. Laćk of correspondence between the interior s pace organization and t h e a r r angement o f t h e exteršor wall surfaces. 3. Lack of correspondence between the form of spa• tial units and t h eir exterior counterparts. 3. O>nis, St. Peter, sot
35 V. Goss: IS THERE A PRE-ROMANESOUE STYLE IN ARCHITECTURE? PFRISTII, 2S/1982, (n, 3~SO)
correspond to th e i n terior space units or t h e i n t erior supports (Marmashen; Cathedral at A ni , b oth c . 1000, and numerous other examples).' Another characteristic worth noting, and St . H r i p sime is a p e r f ect example of it, is the narrowness of passages between the space u nits, or b r i efly, apatial discontinuity, w h ich w e w i l l try to show i s another, fif th, characteristic shared by n p~ 0 the architecture of Eastern Christianity and that of the Pre-Romanesque West.' Naturally, it w o ul d b e i d ealistic t o ex>pect the f i ve c ategories arrived at s o f a r t h r o ugh ou r a n alysis t o a ppear in every Pre-Romanesque building. But we w i l l try to demonstrate that the features such as biaxiality, l ack of c o r r espondence between th e o r g anization o f space and mass, lack o f c o r r espondence between the form of s patial units and t h eir external counterparts, h idden syaces and spatial discontinuity ar e f o und i n a considerable number o f a r c h itectural w ot>ks in t h e West between the eighth and the t w elfth century. Let P 'I us first return to the starting point of cur investigation: monuments of early Croatian architecture. As a countenpart of the Ar menian hidden sanctuary, one may list the hidden westwork of a number of early Croatian churches. The use of a western annex contain• i ng often a m a u soleum an d/or g a l lery — t h e l a t t e r reserved for the r uler or h i s to p o f f icers — spread in t he Eastern A d riatic m ost l i k ely i n c o n nection w i t h the Frankish overlordshi ip (c. 800 — c. 870).' A good 4. St. H r i psi>ne, Etch>niadzin, early seventh century example is the hidden western annex of ,the church of envelope diagran> by Kenneth J. Conant [Sirarpie der St. Mary on t h e I s land i n S o lin (Salona; before 976); iVersessian, The Armenians, Tha>nes and Hudson, Ltd., 1969] in ruins, but relatively easy to reconstruct (Figs. 5 — 6). The church consists of an aisled nave which pr obably c arried on its only square bay a d ome, most likely •
' Most of t hese characteristics seem to b e đetectable in the architecture of late Antiquity. A brief look at the ma• terials collected in Anđrć Grabar's Martyriu>n, 2 vols., Pa• ris, 1942 — 1944, I, Figs. 13, 16, 28, 29, 31, 48, 54, 65, 66, 71, 5. Solin, St. Mary on t he I sland, before 976, groundplan 72, 74,77, 78, 84, 96, 97, 99, showing mostly pagan anđ [author] Early Christian structures, would suffice. It has been no• ticeđ by Der Nersessian that some of the complex Armenian solutions derive from this type of architecture (above, note seven). The nature of the similarities and/or differences between this architecture anđ the Chtristian architecture of both the East and the West during the Early Middle A ges seems to be worth f urther study. Although it is no t possible, in this paper, to enter into this vast area of en• quiry, .it seems that the aesthetic .principles of the archi• tecture of late Antiquity. Eastern Christian world, and the Pre-Romanesque West have a lot in common, and that the precise nature of this phenomenon may be worth a t h o• rough reconsideration. ' Although it appears that the westwork should not be seen as a Katserkirche, at least not from t he outset, the terre• strial ruler soon found its way into the iconography of the western massifs. As demonstrated by Carol H e ttz, .the westwork seems to have been originally reserveđ for the lit>ury commemorating the Savior and Hi s Resurrection. The model that inspired the juxtapositi~ o f a w e stwork ( essentially a centralized struoture) anđ a long i t uđinal chtnz body, seems to have been the complex at the Holy Sepulcher m Jerusalem. See Carol Heitz,Les recherches sur les rapports entre l'architecture et la lilurgie a l đpogue carolingienne, Paris, 1963, pp. 77 ff, 91 ff, 106 ff, 121 ff. As opposed to this thesis, a number of German scholars main>tain that the westwork is p r imarily a Kaiserkirche, lO M s o that the frequent dedication of th e westwork to t h e
36 V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (p. 33 — 50)
in harmony with the local tradrition — encased within a square turret. The rounđed apse was equally hidden within a r e ctilinear mass. There are tw o w e stern an• nexes: a trilpartite exonarthex, accommodating the en• trance to the church and th e stairway to t h e g allery, and an aisled endonatthex, originally supporting a gal• lery and serving as th e m a usoleum of Q u een Jelena (đied in 9 76). Although on e i s o b v i ously c o n f r onted I with three đistinct spatial units, the lateral walls dis• play an uninterrupteđ line of , p ilaster-strips indicating II III IIIII \ III III that the b u i lding m c+t l i k ely a p peaređ as on e s o lid I I I I I Ill III I(Il b lock covered by a u n i f orm g able roof, the only p r o• I.III II II l1 IIIIIIII I jection being the tutlret containing the dome, and the apse." Another, better p r eserved example s eems t o confirm ou r h y p othesis about t h e » h i dden« c h aracter of the westwork of St . Mary's. The Savior's Church at Četina in th e D a lmatian H i ghlanđs (c. 900) is an a i s• leless building with a t r e f oil chevet, a t w o-story wes• tern annex and a t a l I , f i ve-story, tapering t o we r in front of it. The westwovk, inserted in 'between the nave and the tower, is not distinguishable from 'the outside and it shared w it h t h e nave, juđging from t h e t r aces which still remain, a common gable roof (Figs. 7 — 8). One may argue t ha t t h e ch a r actevistics re c a lling those found in t h e w a rd s o f E a stern Christian archi• tecture are to be expected in the builđings at the very outskirts of the western world. Could one establish the presence of the same or similar characteristics in some other Pre-Romanesque families' ? 6. Solin, St. Mary on the Island, reconstruction [author] A group of ,small-scale builđings found within the canton Graubiinden in S w itzerland and d ating mostly from the 8th or 9th century is characterized by a rect• angular, boxlike nave accompanied, at its eastern end, by three alpses (Fig. 9). From the exterior, the building presents itself as a n a i sled st ructure, whereas, upon entering it, ane f inđs oneself in a s i m yle oblong, box• like space (St. Martin an d St . M ar y a t D i ssentis; St . Savior should be explained as a consequence af the merg• J ohn alt Must@ir; St. M a r ti n a t Z i l l is; St . M a r ti n a t ing of the cužt af the Savior with the imperial cult. The Pleiv; St. Peter a t M i s t ail). M oreover, the wh o le t r i • two theses are nat mutually exclusive. Heitz allows for the partite sanctuary can be imbedded within the straight rale af the r uler žn the westwork icanagraphy but can• siđers it of a secondary i mportance,whereas the German termination wall (St. Agatha at Dissen,tis, possibly 10th thesis recognizes the importance af the Nturgy of the Sa• or 1 l th century), recalling the Armenian hidden sanctu• vior, emphasizing that the glarification af the Ređeemer is ary. A similar for m i s f ound agalin within the Adriatic i nseparably intertwined w>th the i mpe vial cult. See i n area, in I s t r ia, an d .the I s t r ian examples have been parbicular Alois Fuchs. »Enstehung and Zweakbestimung der Westwerke,«Westjalische Zeitschrift 100, 1950, pp. 227, pointed out as a l i n k b e tween the Eastern Mediterra• 253 — 255, 259 — 274. Also, Fuchs, »Zur Problem der West• nean and the Central AlI:ps." werke,« in Ka rolingische und Ottonische Kunst, ed. An• A number of featutes we have described so far can dreas Alfoldi et als., Wiesbaden, 1957, pp. 109 — 117. be found in Mozarabic architecture. A horse-shoe apse Heitz's reasoning seems to be valid and applicable pri• marily ta the central lands of the Carohngian Empire. In the borderlands of the western world, secondary ramifi• " On Mazarabic architecturein general, see Jose Fernandes catians seem ta have been quite prominent. Thus the west• Arerms, Mozarabic Architecture, Greenwich, 1972, with excel• work šs reserveđ for a person of distination; št served as lent phatagraphs anđ đrawings af the builđings mentioned a burial chamber, or as a real fortress defending the en• in aur discussion. Far the influence af Mazarabic Iiturgy cn trance ta the church. For further discussion of this problem, some architeatural aspects that interest us here, especially with additional literature, see Vladimir Gvozđanavic, »A an spatial điscontinuity, see Arenas, Mozarabic Archićecture, Note on Twa Early Croatian Royal Mausalea,«peristil 18• p. 220 , especiaily pp. 245 — 246, 265 —266. Also, Pu(g i Cada• — 19, 1975 — 1976, pp. 5 — 10. falch, Le premier art roman, pp. 56 — 67,wha campares the '» For more đetašl see Gvozdanovic, »A note on Twa M r l y relatianship between the architecture of th e Pre-Raman• Croatian Royal Mausalea,«pp. 5 — 10. esque Wets and that of th e Orient, to that between the " Louis Gradecky, L'Architecture Ottonienne, Paris, 1958, Mozarabic, Ambrosian, Gallican on ane hand and oriental p p. 156 — 157, fig. 56. For the role of th e Adriatic in t h e liturgies an the other. See also Louis M. O. Duchesne, Chri• transmisman of the form, see Branko Marusic, »Dva spo• stian Worship, trans. M. L. McClure, London, 1931, p. 93. menika srednjevjekovne arhitekture u G uranu kod Vo d• Haw lilturgIical requirements necessitateđ, and resulteđ in, njana,«Starohrvatska prosvjeta,3rd ser., 8— 9, 1963, pp. better visibility has been điscussed by Heitz, Les recher• 121 — 150. ches, 176 — 177, p. 205.
37 V. Goss: IS THERE A PRE-ROMANESOUE STYLE IN ARCHITECTURF.? PFRISTTT. 25/1982 tn 33 50) w ithin a r e c t ilinear shell appears at Sa n C ebrian d e m onuments of A n g lo-Saxon architecture, Yet, i n t h e Mazote (913), San Miguel de Escalada (913), Santiago case of m o r e c o m plex b u i l dings, either w i t h l a t e ral de Penalba (931), Santa Maria de Melque (before 932) porcehs or t r a nseptal ( B r adford-on-Avon, c. 650 — 700 and San Miguel de Celanova (10th century) (Figs. 10, (Fig. 13); SS, Peter and Paul, c. 600 — 650, and St. Pan• 1 1, 12). A d ome w i t hi n a s q u arish m ass i s f o und a t cras, c, 600 — 650 at Canterbury; Deerhurst, c. 600 — 800 Mazote, Penalba, Melque, Celanova and San Tomas de and later; Reculver, c.650 — 700 and later; Romsey, c. las Ollas (first half o f t h e 1 0th c entury). Actually, a 950 — 1000; Stow, c. 950 — 1000 and l ater), one seems whole triconch is h i dden w i t hin t h e r e ctangular east• j ustified to t ažk about s patial discontinuity, w h ile i n ern end of M a zote. I n a d d ition, spatial d iscontinuity, the case of the buižćings decorated externally with th in t he use of r e l atively n a rrow p a ssages šs one o f t h e stepsof masonry one f i nds ~little, if a ny , c o r respon• standing characteristics of Mozarabic architecture." dence between this fine graphic, wall surface decor and S ome of these characteristics are traceable to the V i • the interior o r ganization of t h e b u i lding (Barnaok, c. s igothic period, for example, the hi dden apse of S a n 950 — 1000; Barton-on-Humber, c. 950 — 1000; Corhamp• Frutuoso .de Mon4elius (c. 670), or the discontinuity in ton, c. 950 — 1000; Earl's Bar t~ , c . 9 50 — 1000; Sompt• t he spatial arrangement of Sa n Pedro de l a N ave (c . ing, c. 1050 — 1100; Woolbeding, c. 950 — 1100).'4 687 — 701) and San Pedro de la M ata (c. 672 — 680), but they seem t o b e come m or e p r o m inent i n t h e M o z a• " Pedro de Palol and Max Hirmer, Early Medieval Art in rabic architecture." Spain, New York, 1967, p. 14 ff, Figs. 4, 5, 6. Due to an overall simplicity — m o st o f t h e examp• " Hamld M. and Joan Taylor, Anglo-Saxon Architecture, 2 vels., Cambridge, 1965, I, pp. 13 — 14, 86 —89, 91 —93, 134 — 143, les being aisleless, longitudinal b u i l dings — f e w of 146— 148, 193— 209; II, 503— 509, 520 — 522, 584 — 593. Also, I, o ur charaoteristics w il l b e f o u n d a p p l icaible t o th e pp. 43~7, 52 — 57, 176 — 179, 222 —226; II, 558 — 562, 684 — 685.
7. Cetina, the Churcb of t he S avior, c. 900, groundplan, 8. Cetina, the Churcb of t he Savior, from southvvest scale l:200 [author j [author]
I •
3y
r $ r
r • ' šr
f~
38 V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (p. 33 — 50)
In Great Moravia, where our enquby is marred by a very fragmentary state of the ruins, one should at least single out the Church No. 9 a t M ž ul cice (9th century), a rounded building with a quatrefoil interior, and the Clntrch No. 6 a t t h e same .place, a rotonda wšth opposed apses, thus exemplifying the principle of bi-axiality." One may again ašject t hat s o f a r w e h a v e u sed e xamples from ithe outskirts of E u rope. Could o u r reasoning be successfully applied to the central lands of the Pre-Romanesque Europe, the core of the Car+ lin~an and Ottonian Empire? On one hand, as we w il l t r y t o d e monstrate in a moment, Čarolingian and Ottonian architecture seem to signify a step toward »Romanesque claritye and »s~ cturažizatienc, but on the other, one cannot pass up the fact that there are features common to other Pre• -Romanesque grou@s anđ monuments. The centralized builđings such as the Palatine Chapel at Aachen (796• 5 la — 804) display the ~bi-axiality, the longitudinal axis be• ing emphasized by grouping of the units along one ho• • ' rizontal line: westworž — centralized core — sanctuary (Eig. 14). Also, in a number of the rotonđas there is an obvious disparity between the form of th e m ass and space. A rounded mass at Wurzburg (8th or early 12th century), an elliptical one at Deutz (1002 — 1019) and a polygonal one at Metlach (987 — 993) contain each a space conceived as a series of projections jutting out from the central core," One should also list .such cases as a half~ t a gon ylaced within the westwork at Essen 9. Mustair, abbe» churcb of St. John, c. 800, view of the (1039 — 1051), the apse hidden within the westwork at apses [Hans E. Kubach, Romanesque Architecture, New Mittelzell (westwork date; 1030 — 1048), or the apsidio• York, 1975] les imbedđed in th e t e rmination w all a t H e l mstedt (mid-11th century?) and Deventer (1027 — 1040)." In each
'~ Josef Cibulka, «L'architecture de la Grande-Moravia au IXe siecle š la lumičre des rćcentes dćcouvertes,«L'infor• ntation d'histoire de I'art, 11, 1966, yp. 1 — 32, especially p. 10. Santiago de Penalba, 941 groundpian [Jacques Fontaine, 24 ff. L,art prćrmnam hispanique, 11, La Pierre-qui-Vire, 1975] " Grodecky, L • Architecture Ottonienne, i~. 164 — 167. Note also the elliptical, more lcngitudinal plan of the church at Deutz, 11. Santiago de Penalba, longitudinal section [Fontaine, " Grodecky, L'Architecture Ottonienne, pp. 60 — 62, 91, 109, L'attt prćroman hispanique, 11] 142.
39 V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (p. 33 — 50)
of these cases the Carolingian o r O t t o nian a r chitect seems to oiperate within th e f r amewcrk o f P r e-Roma• nesque aesthetics. Still, we b elieve it i s p o ssible to d e monstrate that the Carolingian, and subsequently, Ottonian architects ž EB~R e lid contribute toward a m o r e »rational«presentation of an architectural program. As an early example one can list the church of St. Riquier at Centula (799). The architect seems t o d i s t inguish bet w een t h ree u n i ts, two transepts of equal size crowned by s p ires and a simple boxli~ke nave in b e tween (Fig. 15). The spires a re meant to e m p hasize the i m p ortance of t h e u n i t s they crown: the entrance to the church and the access to the sanctuary. The f act t h a t t h e t r a n septs-towers are of an equal size and form may lead the viewer into a confusion, as he may assume that they have the same function or meaning within the architectural program, • • • • • although to the ninth-century visitor t hey probably in•
14. Aacben, the Palatine Chapel and the I » zperial Palace,
• • 796 — 804, g rozznd plan [Kzzbaclz, Romanesque Architecture]
12. San Cebrian de Ma-ote, 913, longitadinal section and gronndplan [Fontaine, L'art p reroman hispanique, II]
13. Br adford-on-Avon, St. Lazvrence, c.650 — 700,groundplan [Kubaclz, Romanesque Architecture]
n I t a r
1 I (J
20 • 0 m.
40 V. Goss: IH THERE A PRE-ROMANISQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (p. 33 — 50)
and western end o f t h e b u i lding. Although the effect t t Q» Nć I M F K fo ~ O C '.N rV I A 4 A N 9 c c 8 Xc. t is ambiguous and, as already stated, may confuse the 16. Hildesheim, St. Michael, 1001 — 1031, groundplan and longitudinal secfion [from Gardner's Art Though the Ages, Fifth Edition, (C) 1970 by Harcourt Brace Jovanovich, Inc.; Copyright 1926 by Harcourt Brace Jovanovich, Inci '%4'm' •'st č' j Renewed by Louise Gardnert Copyriht 1936, 1948, (C) 1959 by Brace Harcourt Jovanovich, Inc. Reprinted by permision 15. Centula, St. Riquier, 799, drawing after the lost of the publisher] Chronicle of Hariulf [Paul Petau, De Nithardo, Caroli Nagni nepote ac tota ejusdem Nitharđi prosapia, breve syntagma, Paris, 1613] O 4 ID 0 0 dicateđ that t h e t w o t o w er s b e long actually t o t w o different churches: the church af the Savior (the west• oooooooooo work) and the church of St . R icharius (the east tran• sept)." One ma y c l ai m t h a t S t . R i q u ier an d r e l ated structures đisplay the c h aracteristic of b i -axiality, or more precisely, bi For a detailed discussion of the signigicance of the orga• mzation of St. Riquer, see Heitz, Les recherches, pp. 21• — 28, 78 ff. " The bipolarity is noticeable anđ the effect remains ambi• guous, even though the accents may be of a somewhat dif• ferent shape and size. A similar phenomenon is found in Mozambic arcMtecture,e.g., at Mazote and Penalba. ~ Grođecky, L'Architecture Ottonienne, pp. 24, 81 ff, 96, 196. " Paul Frankl, Die F r u hmittelaterhche und R o m anische Baukunćs, Potsdam, 1926, pp. 74, 97, 117. ~ Grodeoky,L'Architecture Ottonienne, pp. 56— 58. 41 V. Goss: IS THERE A P>RE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (>p. 33 — 50) I I > I Q 0 I ~ I 18 Mi l an, San An>broglo, n>nth eleventh and t u e l f t h 17. Cardona, San Vincente del Castillo, 1020 — 1038 cent«ry, gro«ndplan [fron> Gardner'sArt Through the Ages, [K«bach, Romanesque Arch>>tecture] Fi ftb Editio», (C) 1970 by Harco«rt Brace Jovano»ich, Inc.; Copyright1926 by Harco«rt Brace Jovanovich, Inc. Re>>ev»ed by Lo«ise Gardner; Copy>ight 1936, 1948, (C) 1959 by Brace Harco«rt Jo»ano»ich, Inc. Reprinted by> pem>ission of the p«blisher] Buildings such as St. Michael at Hi ldesheim seem to q ues, Notre-Dame du P o r t a t C l e r m ont F e r r and) o r mark a t u r n ing point. I n o r der t o p r o gress toward a c ompletely neglected (the original form o f C l uny I I I ) . new style, the architect will s t art t o d i scriminate bet• O ttonian architecture is no t t h e o nl y g r oup w i t h i n ~cen the poles both i n t e rm s o f t h ei r f o r m an d s ize which one finds the signs of a new style. An area which and to relate the exterior t o i n t erior. The ambiguaus should be thoroughly reexamined is that o f t h e » F i r st ~estern apse disappears and the vertical expansion of R omanesque Ar t « of the Lom b a r d-Catalon t y pe. I n the western end by the means of towers is contrasted spite of the noticeable tendency to articulate the walls t o a horizontal expansion of t h e e astern end b y t h e by the means of pilaster-strips, blind arches, or arched means of ambulatory and radiating chaipels. The High corbel tables, a more careful analysis reveals that only 3othic solutions represent this m odel at t h e s tage of i n case of some m ajor m o n uments, this decor i s re• ts full perfection, whereas in th e Romanesque, even lated to the structure of the building (San Vincente de I ully developed Ramanesque, there is still a fairly strong Cardona, 1020 — 38 (Fig. 17) ; Ovarra, B osost, V erdun .mphasis on the vertical grouping around the crossing, — all eleventh century — i n C atalonia; San Ambrogio he western vertical being somewhat restricted (Con• i n Milano, ninth, and eleventh an d tw e l f t h century V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (p. 33 — 50) il~ 19. Oviedo, Santa Maria de Naranco, dedicated in 848, exterior f Kubach, Romanesque Archiiteoture] • ' • u • • • •A' • • • '» (Fig. 18); Rivolta d'Adda, c. 1099?, in Lombardy)." In r ural o r m o d est b u i l dings one r a r ely f i nd s t h e c o r • 20, Oviedo, Santa Maria de Naranco, groundplan respondence between the i n terior and exterior b efore [Fontaine, L'art prćroman hispanique, 1] 1100. Could the growing tendency toward the correspon• l onia, of c o m plex p lans such a s t r i conchs (B r u ll , c . dence be related to the introduction of the vaults? The 1047; Gallifa, 11th and 12th c entury; F a bregues, late e xternal deccration does aippear f i rst o n t h ose p a r t s 1 1th century; M o n t najor, f i r s t q u a r ter o f the 1 2 t h of the building which contain vaults. Before anywhere century; Ponts, early 12th century; Tavernolles, c. 1069; else the strips and arches appear on t h e a p se , the there are even some late 12th century examples such thickening of t h e w a l l r e s ulting i n p H aster-strips r e• as Cellers, Er il l l a V a l a n d P o r q u ieres), polyconchs vealing that there is a special structural element to be (Vallanova, l at e 1 1t h c e n t u ry), c r u c i form s o l u t ions f ound inside and also, by t h eir p o sition on t h e w a l l , (Salou, last t h ir d o f t h e 1 1 t h c e n t ury; L let • ida Sant i ndicating the r a dial d i spersion o f t h e t h r ust o f t h e Ruf, c. 1052; Sant Pau de Camp in B a rcelona, c. 1127) a psidal semidome." Slowly, and seemingly wit h m a ny or centralized b u i l dings o f rat h e r g r o t esque p l a ns i • nconsistencies, this system of wall arbiculation is then ( Planes — o f u n c ertain d ate — a co m b i nation o f a introduced to other facadesP But i t seems that rather than insisting on t h e v a ul t a s a d e cisive factor, one " Puig i Cadafalch, Le premier art roman, pp. 76 — 77, 78, 80, should, place an emphasis on the internal compartmen• 86. Conant, Carolingian and Romanesque Architecture, pp. talization as a guide toward the relating of the exterior 242 — 244, 249. Arthur K. Porter, Lombard Architecture, 4 and interior. The ch urch a t M o n t buš (c. 1032) in Ca• vols., New Haven, 1917, II, p. 532 ff; III, p. 325 . Focillion, Art of the West, I, p. 29, has noticed that a number of the talonia, covered by a co ntinuous barrel vault, displays »First Roma • nesquea bubdžngs failed to b e t r uly Roman• no decoration on its lateral facađesP The introduction esque in terms of their struc • ture. of transverse arches, defining with more p r ecision the u San Vicmuo in Pmto čn Milan, Agliate, Burgal, Estamariu, i nterior units, or ~bays, leads graduallly toward a m o r e St. Ceoilia de Monserrat, to list only a few examples. See Puig i Cadafalch, Le premier art r o ~ a n, pl. 1 0, 12, 13; a nđ more exact pr ojection of t h e s pace units on t h e Conant, Carolingian and Romanesque Architecture, pl. 28A, w all surfaces, a process to f u lly t r i u mp h b y t h e e n d B; 29B. of the eleventh century. A general hesitation within the '» Tables in Puig š Cadafalch, Le premier art roman, pp. 53, iderlined Iby the survival 73, 77, 100, 112. » First Romanesque Art« is u n '~ On Santa Maria de Tolsa de Montbui, see Eduardo Ju• of such » Pre-Romanesquec features as opposed apses nyent, Catalogne Romane, 2 vels., La Pierre~ui-Vire, 1960, (Burgal, 10th to 1 2th century), or, especially in C a ta• I, p. 48 ff. 43 V. t oss:IS THERE A PRE.RGMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (ip. 33 — 50) rotonda and t r i angular p l an) o r d i s p laying t h e » P r e• -Romanesquec lack o f co r r espondence (Cervera, late 1 1th century), rotonda w it h q u a t refoil i n t erior." T h e 1 1th century a r t s h o ul d b e t h e r efore d e scribed a s truly t r a nsitional, th e s u r v ivals o f an e a r l ier s t y l e e xisting side by side w it h t h e elements of a n e w o n e w hich wa s t o co m e t o a f ull b lo s som a r o und t h e year 1100. Beatung in m in d t h i s c h r onological co-existence Pre-Romanesque anđ Romanesque aesthetics in the 11th century, one may l egitimately ask th e f o l lowing q uestion: Could on e l ook b a ckwards beyond th e b e • ginning of the l ith century.in search of the anticipation of the Romanesque style, anđ, conversely, forward into the 12th century, looking for th e survivals of the Pre• -Romanesque'? The answer in b ot h c ases seems to b e aff ivmative. In the 9th century in Asturias, some of the buildings of the » E stilo Ramirense«(843 — 850) đisplay characte• ristics which enable one t o c l a ssify t hem a s a l m ost completely Romanesque.~ Strme of these characteristics seem to have been, present in Asturias at an even ear1i• er date and are adumbrated by San Julian đe los Prados in Gviedo (812 — 842) in terms of corresponden• ce, though still t entative, of mner and outer supports on the chevet and the side walls of the nave. The cor• r espondence is f o r a l l pr a c tical , p urposes perfect a t Santa Maria đe Nara+co (deđicated in 848, Figs. 19 — 20). H ere the space units correspond to t h e i n t erior w a l l surface and vault u n i ts, defined by t r a nsverse arches anđ twisted columns of the wall a r cađes. The interior u nits are then also reflected in the organization of th e exterior wall surfaces defined by projecting buttresses. The only »not-yet-truly-Romanesque« f eature w i t h i n 21. San Salvador de Valdedios, c.893, groundplan this otherwise perfectly logical structure is the a bsencet [Fontaine, L'art prćroman hispanique, I] of a direct structural and visual continuity between the attached columns o f t h e w a l l a r c ade anđ the t r ans• verse arches of the vaults. Sirnilarconcern for correspondence is đisplayed by the remains of the church of San Miguel đe Lillo (848) ; the buttresses corresponđ to the interior supports and the bays to t h e u n it s o f t h e s id e w al l s u r faces. The 22. Solin, SS. Peter and Moses, c. 1070, destroyed, third building oftenconsidered as »Ramirense«, Santa groundplan [authorj Christina de Lma, is less successful in relating the but• tresses to the interior organization, an đ it also » r elap• ses« into th e P r e-Roxnanesque in t e rm s o f a more obvious spatial điscontinuity.~ This may i ndeeđ be an a rgument for s o metimes suspected later d ate o f t h e church, since other post-Ramirense builđings of Astu• rias, such as San Salvador de V aldedios (c. 893), San Salvador de P r iesca (c. 921) and r e lated m o numents announce a similar tendency of reverting to the Pre-Ro• manesque aesthetics. At Valdedios (Fig. 21) there is no corresponđence between the arrangement of the inner " Junyent, đatalogne Romane, I, pp. 32, 34, 35, 36, 37, 39, 59; II,28 — 29, 31— 32, 47 ff. Jose Puig i Ca đafalch, L'Arqui• tectura romanica a Catalunya, Barcelona, 1911, pp. 318, 321• — 324. Marcel Durliat, Roussillion Roman, La Pierre-qui-Vire, 1958, pp. 24 — 25. u Already noted by Conant, Carolingian and Romanesque Architect+re, p. 42 ff. " On all these buildžngs see Canant, Carolingian and Ro• IO M manesque Architecture, pp. 42 — 45. V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (ip. 33 — 50) 23. Zadar, St. Krsevan, c. I175, south mal/ [author] space and the exterior w all s u r faces, in ad d i t i on to surfaces (a somewhat similar f eature is d i splayed by which one also discovers a » h idden« w estwork.-" the side wa~lis of Santiago de Penalba), the church at The early yrecocity of t h e » Ramirense«group is only Sady in Great Moravia (9th century), .if one is to t r u st rerely matched b y b u i l d ings w i t hi n o t her P r e-Roma• the reconstruction on the basis of very scanty remains, nesque schools. Among th e M o zarabic b u i ldings one and the eleventh century church of SS. Peter and Mo should isolate Santa Maria de Bamba in terms of cor• ses in Solin in Croatia (c. 1070; Fig. 22)." The ruins of r espondence between the b ays and t h e e x terior w a l l this large aisled church w i t h a s o m ewhat n a r r ower westwork display a,perfect correspondence between the bays defined also by pilaster-strips. In the interior one ~ On San Salvador de Valdedios and other buildings of the period of Alfonso II (866 — 910), see A. Bonet Correa, Spanish can easily imagine transverse arches bringing about an pre-Rornanesque Architecture, Greenwich, 1967, pp. 174 e ven more .perfect and m ore y r ecise idefinition of t h e — 211. b ays, although t h ere i s n o d i r ec t e v i dence that t h e " Arenas, Mozarabic Architecture, pp. 8 4, 110; Cibulka, »L'architecture,«12 — 13; Dyggve, History, p. 133. building was vaulteđ. The hidden sanotuary, consistmg of a square apse and rounded apsidioles, all imbedđed into the straight termination wall, .is, however, still a Pre-Romanesque feature. There is another early Croa• 24.Zadar, St. Krsevan, gounćplan [author] tian building w h ich m a y b e m e n t ioned here — the chapel of St. Nicholas on the Island of Lopud (c. 1100). This church ~belongs to the same .type as St. Peter at Omis, the building w it h w h ic h w e h ave i n i tiated our discussion, but w h a t m a k es .it u n i que a m ong s o me thirty sister monuments is the fact that th e t r i partite organizationo f the nave (with the dome on top of t h e : : • : : • : : Q : : • : : Q : central bay) corresponđs to an equally t r i partite orga• nization of th e l ateral facades. Keetping in imind what we have said about the analytical tendencies in Ottonian architecture, one may concluđe that the » Romanesquec characteristics exist simultaneously with the »Pre-Ro• manesque«ones, and in some cases can ~betraced back II II as far a s t h e 9t h c e n tury. Th e f r e q uency i n c r eases IIII throughout the l l t h c e n tury an d th e Romanesque be• c omes a f u ll y f o r med s t yle a r ound 1 100 w i t h su c h buildings as Cluny I I I , S t . S ernin i n T o u l ause, Santi• IO M ago de Compostella, San Ambrogio in M i l ano. V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE2 PERISTIL 25/1982. (p. 33 — 50) 25. Perigueux, ht. Etienne, begun c. 1/00, interior, the east bay [Kubach, Romanesque Architecture). O ne should b e g u arded against p r oclaiming t hi s a in the l a te r 1 2t h c e n tury; M o n t agrier, R o manesque final, definite victory o f t h e R o m anesque style. After building of u n d ertermined date). The same area d i s• the year 1100 the majority of k ey, style-promoting mo• play also interest in some oth er » s t rangee plans, wit• numents, and the buildings depending on them, display ness the octagon-octaconch of St. M i chel d'Entraygues Romanesque characteristics as we have seen them de• (c. 1137), or the Greek cross churches at M agnac-sur• fined in the first part of t his stuđy. But there are still -Touvre (12th century) an d M a i n fonđs ( 1160 — 1180)." a number of f ine monuments, or entire groups, which O ne should also note the stubborn persistence of t h e could not b e c o n sidered t o tally R o manesque. T h ose rotonda church i n t he Ro m a nesque H u ngary ( s ome monuments or groups often somehow left o u tside the eighty examples, some of w h i c h d i splay th e a b sence main stream of the research in the art of t h e eleventh of correspondence between the interior and exterior), anđ twelfth century, or j u s t b r i efly commenteđ upon, and, at yet another end of the western world, in Scan• deserve special a t tention. T hey r e p resent d e v i ations dinavia." Sometimes one encounters b u i ldings w h i ch from what is considered the standard model and may illustrate some specific cultural situation wi thin a cer• tain milicu. This area of t h e » A nti-Romanesque«seems "' On Worms and Laach, Frankl, Fruhntittelalterhehe und to be a f e r t ile f ield fo r f u t ur e r esearch, and here we Rotnanische Baukunsl, pp. 189 ff, 249, 270. On the Cologne group, ibid., pp. 89 ff, 269 — 271. On Catalonia, see above note s hall just t r y t o s k e tch a f e w li n e s o f s u c h a n e n • twenty-eight. On Southwestern France, or Ocoitania, Jacques quiry. Brosse et als., eđs., Dictionnaire des eglises de France, 5 vols., As an obvious example of t h e P re-Romanesque tra• Paris, 1966, II, pp. 2 C 152; III, 3 8 10 — 11, 3 8 155 — 156, 3 8 166, 3 8 79, 3 8 103, 3 C 106 — 107, 3 C 96 — 97, 3 C 93 — 94, dition one ma y l i s t t h e poipularity o f the bi- .polar 3 C 182 — 183. Also, Jean George, Les eghses de France• model (Worms, M aria L aach) and o f s o m e c o m plex, Charante, Paris, 1933, p. 150. primarily ,polyconch plans (the Cologne group) in Ger• " On H ungarian rotondas, Veronika Gervers-Molnar, A man Romanesque. Another group of triconchs is found kozepkori Magyar-or-szag rotundai, Bu dapest, 1972. The i n the French Southwest (Aubiac and G ueyze in t h e author lists eighty rotondas anđ t e n o t her c e ntralized churches đating from the period between the 10th and 13th late 11th century; St. Martin-de-Lonđres and Tourtoirac centuries. On Scandinavia, Aron Andersson, The A rt o f in the early 12th century; Montmoreau and St. Macaire Scandinavia, 2 vola., London, 1970 II, pp. 141 — 147. V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/'1982. (p. 33 — 50) appear to b e » p erfectly Romanesque,«yet they indeed Within this ccntext we should like to b r i efly comment a re not. The exterior w al l t w elve units arcade of t h e on one c h aracteristic o c citan c r eation: t h e »d o m eđ churcb of St. K r sevan in Zadar (c. 1175) in Croatia is =hurches of Aquitaine.e in no way related t o the seven bay organization of t h e interior, thus showing the persistence of Pre-Romanes• The absence of an y c o nsiderable number o f a i s leđ que aesthetics in t h e m i n d o f C r o a tian R o manesque lvuilđings in the architecture of the French South-West architects (Figs. 23 — 24)." ivas noticed long ago."' The t r aditional aisleless chur• This conservativism may b e d u e t o t h e p e r i pheral ches, sometimes of r emarkable đimensions, started to position of t h e c o u ntry ( H u ngary, Croatia, Scandina• be covered, around 1100, atleast in a number of cases, via), or to a desire to perpetuate a certain distinguished by a series of domes. Within the group of c. 80 domed l ocal tradition, such as a n i r rsperial t r adition i n G e r • churches one can easily i dentify th e m a j or , s t yle-for• many. However, in case of South-Western France-Occi• ming monuments from a m a ss of e ssentially modest, tania it seems that a certain conservativism in c losely r ural b u i lđings vulgarizing th e f o r m s of the gre a t intertwined w it h a n o ther aspect, openness to f o r eign mođels.~ The cruciform, five-domed St. Front i n Pćri• cultural,inf lux. Thus some extraordinary forms created gueux apart (c. 1120), one seems justified in s peaking i n the South-West may b e seen i n a s t r a nge way a s a bout basically tw o g r o ups. One centers around t h e motivated both by conservativism and cosmopolitanism. Cathedrals o f C a h ors ( c . 1 100 — 1125) and P ćrigueux ( c. 1100 and l ater), another ar ound th e C a thedral of 26. Pćrigueux, St. Etie>rne, groundplan [Jean Secret, Angouleme (c. 1110 — 1130) (Figs. 25, 26, 27). In the case Pćrigord Roman, La Pierre-qui-Vire, 1968] of the first g roup (rural examples excepted) the domes were originally visible from t h e o u t side." I n t h e c ase of the second group the đomes were from t h e o u tset masked by a gable roof." The first grou>p is»Romanes• >I I >I> Ir >I > I > Dr. Gervers-Molnar draws my attention to the fact that I> the rotonda form is so frequent in Eastern Central Europe >I that it should be considered the regular form of Roman• II I I > II csque architectural expression. ,She also points out to me I I > that my definiticn of the Rort!anesque is very narrow, with which I totally agree. The fact, however, remains that in the countries such as France, where builđings close in form to my»ideal modele occur with a high degree of frequency, the other forms, such as rotonda, represent a>n accident rather than a rule. I am in no way claiming that my system is perfect; it is only an attempt to contribute to what I I I I believe to be a fruitful area of research and discussion. I/ X l > 10 -' Ciril Ivekovic, San>ostan i crkva Sv. Krsevana u Zadru, Zagreb, 1931. Another peripheml country, this time of t he Byzantine sphere, should be a>t least briefly m e ntioned ( h ere. This is Serbia. Numerous Serbian lžth a n d 13t h I l century buildings display a keen šrrterest in»Romanesque« r I decorative rnotifs. However, in most of the cases there is very little correspondence, between the form o f i n t erior spaces anđ exterior mass. For example, in the case of the church at Sopocani (c. 1235) an inscribed cross spatial so• luticn is hidden within what appears to be arr aisled ba• silican shell! In a n excellent paper given at th e Annual Meetings of the Society of the Architectural Historians in San Antonio, April, 1978, Professor Slobodan Curcic has applied a system similar to m ine >in assessing the amount of the Romanesque influence in Serbian međieval churches. I am grateful to Professor Curcic for his reađiness to com• ~ XI ~ eS i eele pare bis notes with mine. -' Eugene Lefćvre-Pontalis, »L'ecole du Pćrigorđ n'existe II il gg XII — Sieele pas,e Bulletin Mr>nu>nental 82, 1923, pp, 7 — 35, especially I pp. 25 — 35. "" Here is a brief list of the most sigr>iHcant contributeons Porčies đisparrtes lh p I to the study of the »domed churches of Aquitaine«: Charles r de Verneilh, L'Architecture byzantine en France, Paris, 1851; Raymond Rey,La Cathedrale de Cahors, Paris, 1925; Joseph Roux, La basilique St. Front de P ćrigueux, 1920, Marcel Aubert, »Lesćglises romanes du Pćrigord,«ćongres Arche• logique 90,1927, pp. 392— 401.Some luciđ remarks have been made by Francastel, L'Hu>nanism roman, pp. 35 — 36, 138• — 139, 176 — 178, 215. »' Le Marquis de la Fayolle, »ćgžise de Grand-Brassac,e Con• grčs Archeologique, 30, 1927, pp. 363 — 375. l • r . • «' T U p n n • n n b n A L š • L TE « n • iL • T v l n YLT • np • • • T n / T v [ nr « v, voss • žo l šl c A c 6 r A c Av l«N C«c o + U c o l I Lc ID I A A \ I l l ! c ć l U žćc r PERISTIL 25/1982. (tp. 33 — Sv) que« in the frankness with which it displays its unusual between the space and mass. The interior is a i sleless, type of nave vault and roof, but the domes make them consisting of a series of square, cubic, domed bays, The appear exotic a n d »u n -Romanesque.« In t e r m s o f a same lack of i n t e rest i n c o r respondence is d i splayed tnechanical repetition of squarish, domed units wh ich, by the architectural decoration of t h e l a teral facades. as it were, could be m u l t iplied ad i n f in itunt, the buil• Whereas the position of the bays is indicated by projec• dings exude a strange, oriental f lavor. The use of stra• ting buttresses, the surface in between the bu t tresses ight lines, fIlat chevets, cubic masses and spaces (espe• is decorated by two r oundheaded arches as opposed to cially St. Etienne at Pćrigueux; St. Avit Seignieur, early a three arch sequence in t h e i n t erior." S i m t larly, at 12th century; Tremolat, Cherval, Bourg«đes-Maisons, all Fontevraud (deđicated in 1119), the side walls buttres• 12th century), th e b r oad an đ t al I .pointed arches on s es of the m o nastery churcb announce some kind o f the exterior walls (which appear also in churces covered alternating system, whcreas, in fact, one discovers an by a more traditional barrel vault, e. g., St. Privat-des• interior consisting o f a n a i s leless nave containing a - Prčs, first half of t h e 12th century, as well as i n t h e series of cubic domed bays (Fig. 28)." dtxmed churches, e. g ., Cherval, Paussac, Ver teillac, Allemans and numerous other twelfth century buildings in the vicinity of Pćrigueux) reinforce this oriental ef• f ect. This o r i entalism i s f u r t he r e m p hasized by t h e use of elements traditionally recognized as bor rowed 27. Angoulćme, Cathedral, c. II00 — II30, groundplan [Charles Daras, Angoumois Roman, La Pierre-qui-Vire, 196I] from the M u slim w o r l d p o lylobed arches and nt o dil• lonsw-copaux. In a n u m ber o f c a ses th e s pace i s l o n gituđinally oriented by an apse (Cahors), or an apse accompanied b y radiating chapels (Souillac, first h al f o f t h e 1 2 t h c entury), or by a s q uare mass larger in size than t h e other u nits (both the original and definite from of S t . Etienne at Pćrigueux) (Figs. 25, 26). But i n a n u m b er of buildings (St. A vi t S e ignieur, Cherval, B o urg-des• -Maisons) this sense of đ i r ection i s f o r a l l pra c t ical pu@poses non-existent. Anđ whereas the already men• tioned blind arches, often pointed, but also roundhea• d eđ, correspond in m any cases to the d i vision of t h e interior space (Cherval, Paussac), this need not always be the case as witnessed by the older of t h e tw o p r e• served bays of St. Etienne at Pćrigueux. Here the late• ral facades are decorated by two blind arches, each unrelated to the single, large wall a~ i n t h e i n t erior."' A different, seemingly less radical sotution is offered by the Cathedral of Angouleme and its progeny (Fig. 27). Here the exotic elements, the domes, are hidden under• n eath a c o n t inuous r oof, an d t h e s i l houette, wit h a transept and radiating chapels, is perfectly » R omanes• que.«The facade announces to the visitor a t r i p artite, a isled ,interior. B u t up o n e n t e ring t h e c h u r ch, o n e immediately notices the absence of any correspondence Renć Crozet,»Remarques sur le repartition des ćglises h file des coupoles,«Cahiers de civilisaiion Inedievale, 4, 1961, 175 — 178. l» Gn the Romanesque architecture in Pćrigor đ in general, see Jean Secret,Pćrigord roman, La Pierre-qut-Vire, 1968. " On Angoulćme Catheđral, Charles Daras, Angounlois ro• man, La Pierre-qui-Vire,i 1961, pp. 69 — 94. C • " On Fontevrauđ, Pierre đlBrbćcourt anđ Jean Porcher, Anjou roman, La Pierre-qui-Vire, 1959, pp. 17 — 31. ~ This characteristic is most noticeable at St. Front at Pćrigueux, and the cathedrals at Pćrigueux (eastern bay), Cahors and Angoulčme. Most of the rural churches, howe• I: ver, do not take advantage of this feature. There is usually X II •ič just one small winđow per bay anđ the naves remain daržc. Since some of the domes seemeđ to have been coveređ by X»iii • •, a roof from .the outset, anđ others have only a fem (up to four) small windows, the dome does not serve as a lantern. Lx,3 X I X i • The light comes through the windoćvs šn the siđe walls and enters the space obliquely, concentrating thus on the lower and intermeđiary tiers of the interior space. 48 V, Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (p. 33 — 50) 28. Fontevra»d, abbey ch»rcl>, dedicated in ll / 9 , in terior ton • ard the east [K>rbacI>, Romanesque Architecture]. Even on the basis of such a brief discussion as ours, St. Denis would have certainly commended the optical one may conclude that t h e » A nti-Romanesque« tenden• quality of the interiors of St. Front or the Cathedral of cies vie quite successfully wit h th e R o manesque ones A ngouleme. And a f ter a l l , t h e tr a đ i t ion o f dom e d , in the twelfth century art o f O c citania anđ that some aisleless churches survived đirectly i nt o t h e A n gevin of the m ost ty>pical and m ost a t t r active creations of Gothic. Aquitanian art s i gnify a d e parture f ro m R o m anesque A t the end o f o u r br i e f e n q uiry, w h ich s h ould b e logic and clarity. In one respect, however, the Occitan continued and amended through research in many areas » irrationalisme seems to p o int t o w ard th e G o thic, I n and aspects briefly điscussed above, one seems justified the domed churches the thrust is naturally assumed by in suggesting, at l e ast t e ntatively, the f o l lowing con• four supports in the corners of the bays. This enabled clusions: the architect t o o p e n u p t h e w a l l , a s i n By z a ntinc 1. There seem to be architectural features which per• architecture (e. g., Hagia Sophia and H agia E i rene in mit one to stylistically distinguish between the Pre• Constantinople), into a series of windows. Romanesquc -Romanesque and Romanesque. architecture, once the gorin vault was perfected, deve• 2 . Those features, identified at t h e b e ginning of o u r loped essentially th e s ame »four point~ system, and stuđy, Bre: prepared the way for the elimination of th e stt~ctural a. Bi-axiality or b i . polarity i n t h e P re-Romanesque role of the wall šn the Gothic. Yet there is hardly any as opposeđ to m o no-axiality an d m o no-polarity other early twelfth century vaulteđ church receiving as in the Romanesque. much light as St. Front anđ his sister builđings, since b . The a b sence o f cor r e spondence b e tween t h e the architect was able to open up almost the whole sur• organization of s p ace and t ha t o f t h e e x t erior f ace of t h e w a l l a r c hes underneath th e d o mes i n t o wall-surfaces. v vinđows which, in a d d i tion, b r ing t h e e x terior l i g ht c. The lack of correspondettce in form between the into an aisleless, fairly unified space.~ Abbot Suger of space units and their external shells. 49 V. Goss: IS THERE A iPRE ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (y. 33 — 50) d . The ipresence of h i dden i n terior u n it s u nđistin• monuiments of t h e e l eventh centur y » F i rs t R o m a• guishable from the outside. nesque Art« a r e i n f a c t s t il l P r e-Romanesque. The e. Spatial discontinuity as o pposed to syatial con• eleventh century is an age of transition. tinuity. 6. The change from the Pre-Romanesque to Romanes• que, from the Early to the High Middle Ages, should 3. Many o f t h ese characteristics ar e s h a red b y all be seen as a long p r ocess, taking about t h ree cen• groups of Pre-Romanesque architecture in the West, turies, during which period various, often different at least between the 8th and the l l t h c e n tury, anđ stylistic tenđencies, currents and undercurrents exist one seems justified in using the term Pre-Romanes• side by side. que to c o ver s u ch s eemingly đisparate groups of The approach we are proposing here obviously means architectural m o n u ments a s A s t u r ian, M o zarabic, c omplication ~rather than simplification of a n ol d p r o • Carolingian, Early Croatian, etc. It i s w o r t h n o ti ng blem. But b y m a k ing th e p i c ture m or e c o m plex one that similar characteristics are found i n t h e a r chi• also hopes to make it r i cher, more tr u t hful, less regu• tecture of Eastern Christianity and t hat one m i ght l ar, and t h erefore m or e h u m an; w h a t e m e rges i s a speaka bout t h e » orientajtzing~ character o f . t h e Pre-Romanesque. However, this particular question picture of a s l ow , g r adual and t o r t uous transition of the Western Međieval World from its chilđhood toward will require careful f u r t her i nvestigation. Also, one the age of youth anđ maturity. m ay wish t o i n v estigate to w h a t e x t ent al l t h o se c haracteristics, be i t i n Ea s t ern C h r i stian o r P r e • - Romanesque architecture, depenđ on c e r tain l a t e Antique forms. " I am g rateful to Dr. Gervers-Molnar far twa additional remarks .of thearetical nature. First of all, she notes that the 4. Romanesque architecture đisplays consistently cha• Romanesque should be seen as a result of the economic racteristics opposite to t h ose of t h e P r e-Romanes• growth in the eleventh and twelfth centuries. I fully agree q ue. The tw elfth c entury d eviations from t h e R o • that the economic factor should be taken into consideratian in a braader study attempting to place architecture within manesque model seem to be another area of fruitful the cultural history af th e period. She also reminds me research. that ane of the key differences between the fully developed 5 . Romanesque tenđencies can 'be traced back t o t h e R amanesque and e arller m eđteval architecture i s the appearance of the facade sculpture in the former. When I ninth century. They grow i n f r e quency throughout wrote this anttcle I did nat plan ta enter into a discussion the eleventh century, t ake o ver a r o und 1 100, and of that question, although, recently, I dealt with it at least continue into the Gothic. One may actually maintain in a preparatory manner in my paper an »Art As Human that the Romanesque achieves,its total f u l f i llment Experience: Remarks on the Rise of the Monumental Style in the Middle Ages,«presented at the 1978 Annual Meetings in the absolute elarity of the Hihg Gothic structure, of the College Art Associatian. Once more, I would like to t he monoaxiality of th e H igh Gothic space and t h e express my gratitude ta Dr. Gervers-Malnar far her useful supreme continuity of the Gothic interior. Numerous remarks. S ažet a k POSTOJI LI U ARHITEKTURI PREDROIVIANIČKI STIL? Poznato je, nazivi su m edijevalnih stilova umjetnosti po• Ta arhitektura znači korak prema .romaničkoj čistoći. i nešto proizvoljni. No, dak je više-manje razumljivo što mislimo »strukturi., ali tu ipak postoje oblici koji su zajednički predra• pod nazivom romanika i gotika, kao da nema slaganja u tome maničkoj grupi, jer niz primjera dokazuje da su još uvijek unu• što je predromanika. Istražujući bitne značajke romaničkog i tar okvira predromaničke estetike, premda se ne može poreći predromanićkog izraza, ovaj prilog nastoji odgovoriti na ta pi• da je njihov arhitekt uspješno analizirao cjelinu, te odvojio je• tanja u primjerima arhitekture između 800 i 1100 kršćanskog dinice različitih funkcija i važnosti. Zapada. Imajući na umu da u jedanaestom stoljeću koegzistiraju ro• Predromaniku označava: manika i predromanika, mogli bismo se upitati treba li roma• 1. Biaksijalnost. Naglašenoj ulozi vertikalne osi u formiranju nički stil tražiti već početkom ili čak i prije jedanaestog sto• mase suprotstavlja se longitudinalna. os prostorne organi• ljeća, odnosno preživjelu predromaniku još u dvanaestom sto• zacije; ljeću? Vrijedi jedno i drugo, što se opširno dokazuje na pri• 2. Ne postoji odnos između unutrašnje prostorne organizacije i mjerima Asturije, Hrvatske, rajnskog područja, francuskog ju• vanjskih zidnih površina; gozapada, Mađarske i Skandinavije. .Konzervatizam. se može 3. Nema odnosa između oblika prostornih jedinica i njihove pripisivati periferalnom području (Mađarska,Hrvatska, Skandi• vanjske pratnje; navija) ili želji da se ponavlja stanovita lokalistička tradicija, 4. Postoje skroviti prostori na čije postojanje ne ukazuje vanj• all i k o zmopolitizam«, što napose vrijedi za jugozapad Fran• ska analiza; cuske. Može se zakljućiti iz analize niza primjera da se »anti• 5. Nailazimo na prostorni diskontinuitet. romaničke tendencije uspješno takmi če sa romaničkima u Neki se od tih kriterija nalaze već u arhitekturi istočnog umjetnosti Okcitanije i da se najatraktivniji primjeri Akvitanije ranokršćanstva. okreću od romaničke logike i jasnoće, a u neku ruku okcitan• Autor niže i analizira primjere iz Dalmacije, Švicarske, Špa• ski iracionalizam vodi već i prema gotici. njolske, Engleske, Maravske — a mogu li se ti kriteriji primi• jeniti i na predromaniku središnje Evrope, jezgru karoI'inškog i Na kraju istraživanja mogu se barem probno sugerirati ovi otonskag carstva? zaključci: 50 V. Goss: IS THERE A iPRE ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (y. 33 — 50) d . The ipresence of h i dden i n terior u n it s u nđistin• monuiments of t h e e l eventh centur y » F i rs t R o m a• guishable from the outside. nesque Art« a r e i n f a c t s t il l P r e-Romanesque. The e. Spatial discontinuity as o pposed to syatial con• eleventh century is an age of transition. tinuity. 6. The change from the Pre-Romanesque to Romanes• que, from the Early to the High Middle Ages, should 3. Many o f t h ese characteristics ar e s h a red b y all be seen as a long p r ocess, taking about t h ree cen• groups of Pre-Romanesque architecture in the West, turies, during which period various, often different at least between the 8th and the l l t h c e n tury, anđ stylistic tenđencies, currents and undercurrents exist one seems justified in using the term Pre-Romanes• side by side. que to c o ver s u ch s eemingly đisparate groups of The approach we are proposing here obviously means architectural m o n uments a s A s t u r ian, M o zarabic, c omplication ~rather than simplification of a n ol d p r o • Carolingian, Early Croatian, etc. It i s w o r t h n o ti ng blem. But b y m a k ing th e p i c ture m or e c o m plex one that similar characteristics are found i n t h e a r chi• also hopes to make it r i cher, more tr u t hful, less regu• tecture of Eastern Christianity and t hat one m i ght l ar, and t h erefore m or e h u m an; w h a t e m e rges i s a speaka bout t h e » orientajtzing~ character o f . t h e Pre-Romanesque. However, this particular question picture of a s l ow , g r adual and t o r t uous transition of the Western Međieval World from its chilđhood toward will require careful f u r t her i nvestigation. Also, one the age of youth anđ maturity. m ay wish t o i n v estigate to w h a t e x tent al l t h o se c haracteristics, be i t i n Ea s t ern C h r i stian o r P r e • - Romanesque architecture, depenđ on c e r tain l a t e Antique forms. " I am g rateful to Dr. Gervers-Molnar for two additional remarks .of theoretical nature. First of all, she notes that the 4. Romanesque architecture đisplays consistently cha• Romanesque should be seen as a result of the economic racteristics opposite to t h ose of t h e P r e-Romanes• growth in the eleventh and twelfth centuries. I fully agree q ue. The tw elfth c entury d eviations from t h e R o • that the economic factor should be taken into consideration in a broader study attempting to place architecture within manesque model seem to be another area of fruitful the cultural history of th e period. She also reminds me research. that one of the key differences between the fully developed 5 . Romanesque tenđencies can 'be traced back t o t h e R omanesque and e arller m eđteval architecture i s the appearance of the facade sculpture in the former. When I ninth century. They grow i n f r e quency throughout wrote this anttcle I did not plan to enter into a discussion the eleventh century, t ake o ver a r o und 1 100, and of that question, although, recently, I dealt with it at least continue into the Gothic. One may actually maintain in a preparatory manner in my paper on »Art As Human that the Romanesque achieves,its total f u l f i llment Experience: Remarks on the Rise of the Monumental Style in the Middle Ages,«presented at the 1978 Annual Meetings in the absolute elarity of the Hihg Gothic structure, of the College Art Association. Once more, I would like to t he monoaxiality of th e H igh Gothic space and t h e express my gratitude to Dr. Gervers-Molnar for her useful supreme continuity of the Gothic interior. Numerous remarks. S ažet a k POSTOJI LI U ARHITEKTURI PREDROIVIANIČKI STIL? Poznato je, nazivi su m edijevalnih stilova umjetnosti po• Ta arhitektura znači Imrak prema .romaničkoj čistoći. i nešto proizvoljni. No, dak je više-manje razumljivo što mislimo »strukturi., ali tu ipak postoje oblici koji su zajednički predra• pod nazivom romanika i gotika, kao da nema slaganja u tome maničkoj grupi, jer niz primjera dokazuje da su još uvijek unu• što je predromanika. Istražujući bitne značajke romaničkog i tar okvira predromaničke estetike, premda se ne može poreći predromanićkog izraza, ovaj prilog nastoji odgovoriti na ta pi• da je njihov arhitekt uspješno analizirao cjelinu, te odvojio je• tanja u primjerima arhitekture između 800 i 1100 kršćanskog dinice različitih funkcija i važnosti. Zapada. Imajućl na umu da u jedanaestom stoljeću koegzistiraju ro• Predromaniku označava: manika i predromanika, mogli bismo se upitati treba li roma• 1. Biaksijalnost. Naglašenoj ulozi vertikalne osi u formiranju nički stil tražiti već početkom ili čak i prije jedanaestog sto• mase suprotstavlja se longitudinalna. os prostorne organi• ljeća, odnosno preživjelu predromaniku još u dvanaestom sto• zacije; ljeću? Vrijedi jedno i drugo, što se opširno dokazuje na pri• 2. Ne postoji odnos između unutrašnje prostorne organizacije i mjerima Asturije, Hrvatske, rajnskog područja, francuskog ju• vanjskih zidnih površina; gozapada, Mađarske i Skandinavije. .Konzervatizam. se može 3. Nema odnosa između oblika prostornih jedinica i njihove pripisivati periferalnom području (Mađarska,Hrvatska, Skandi• vanjske pratnje; navija) ili želji da se ponavlja stanovita lokalistička tradicija, 4. Postoje skroviti prostori na čije postojanje ne ukazuje vanj• all i k o zmopolitizam«, što napose vrijedi za jugozapad Fran• ska analiza; cuske. Može se zakljućiti iz analize niza primjera da se »anti• 5. Nailazimo na prostorni diskontinuitet. romaničke tendencije uspješno takmi če sa romaničkima u Neki se od tih kriterija nalaze već u arhitekturi istočnog umjetnosti Okcitanije i da se najatraktivniji primjeri Akvitanije ranokršćanstva. okreću od romaničke logike i jasnoće, a u neku ruku okcitan• Autor niže i analizira primjere iz Dalmacije, Švicarske, Špa• ski iracionalizam vodi već i prema gotici. njolske, Engleske, Maravske — a mogu li se ti kriteriji primi• jeniti i na predromaniku središnje Evrope, jezgru karoI'inškog i Na kraju istraživanja mogu se barem probno sugerirati ovi atenskog carstva? zaključci: 50 V. Goss: IS THERE A PRE-ROMANESQUE STYLE IN ARCHITECTURE? PERISTIL 25/1982. (p. 33 — 50) 1. Postoje arhitektonski oblici koji ukazuju na različite stilističke se ta obilježja moraju još pažljivo dalje istraživati, naročito značajke između predromanike i romanike. ako postoje i u kasnowfhknim oblicima. 2. Ti su oblici: 4. Romanička arhitektura pokazuje dosljedno karakteristike su• a) biaksijalnost ili bipolaritet u p redromanici nasuprot mo• protne predromanici. Slučajevi devijacije od romaničkih mo• noaksijalnosti i monopolariteta romanike; dela dvanaestog stoljeća još su jedno polje plodnog istra• b) ne postoji odnos između prostorne organizacije i vanjskih živanja. zidnih površina, dok se u romanici taj odnos strogo pro• 5. Romaničke se tendencije mogu pratiti već u devetom sto• vodi ; ljeću, rastu u jedanaestom, prevladavaju oko 1100 godine c) ne postoji odnos u oblicima prostornih jedinica i njihovih zatim kontinuiraju u gotiku. Može se čak braniti da roma• vanjskih Ijuski, dok se u r o manici taj o dnos redovito nika postiže apsolutni razvoj u čistoći visokogotičke kon• nalazi ; strukcije, u monoaksijalnosti gotičkog prostora i u njegovu d) postoje »potajne«unutrašnje jedinice, koje se n e r aza• superiornom kontinuitetu. biru izvana, na što ne nailazimo u romanici; Mnogobrojni spomenici jedanaestog stoljeća »prve romaničke e) prostorni diskontinuitet suprotstavlja se prostornom kon• umjetnosti « zapravo su predromanika. Jedanaesto stoljeće je tinuitetu. prijelazno. 3. Mnoge navedene značajke posjeduju sve grupe zapadnjačke 6. Prijelaz od predromanike na romaniku, od ranog do visokog predromanike, barem one između osmog i jedanaestog sto• srednjeg vijeka treba promatrati kao proces, dugačak oko ljeća, pa tako taj naziv p r e d r o m a n i k e pokriva naoko tri stoljeća, u kojem su razdoblju postojale različite, često bitno različite grupe asturskih, mozarabičkih, karolinških, suprotne stilističke težnje i tokovi. starohrvatskih arhitektonskih spomenika. Studija ujedno pruža sliku o polaganom, postupnom i vrluda• Slične se značajke mogu naći u ranokršćanskom Orijentu, pa vom prijelazu srednjovjekovnog svijeta od djetinjstva i mladosti se može govoriti i o »orijentalnom«biljegu u predromanici, ali do pune zrelosti. 51