The Labyrinth of Reims Cathedral

Author(s): Robert Branner

Source: Journal of the Society of Architectural Historians, Vol. 21, No. 1 (Mar., 1962), pp. 18-
25
Published by: University of California Press on behalf of the Society of Architectural Historians
Stable URL: http://www.jstor.org/stable/988130 .
Accessed: 04/09/2013 06:50

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

University of California Press and Society of Architectural Historians are collaborating with JSTOR to
digitize, preserve and extend access to Journal of the Society of Architectural Historians.

http://www.jstor.org

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
 The Labyrinth of Reims Cathedral

ROBERT BRANNER Columbia University

LAID out like a great emblem on the floor of the nave, the
famous lost labyrinth of Reims cathedral once commem-
orated the first architects of the Gothic monument (fig.
1).1 There were images of masons in the four corners of
the maze, with inscriptions stating their names, their
lengths of service and the parts of the edifice they had
worked on, and another image, which will now probably
never be precisely identified, in the center. The work un-
fortunately was destroyed in 1778 and is known to us only
through copies.
Throughout the nineteenth century the labyrinth was
generally treated as a piece of curiosa in guidebooks and
studies of the cathedral.2 It was Louis Demaison who first
showed, in 1894, that the masters named in it were the
veritable thirteenth-century architects of the cathedral.3
He listed all the known copies, including the oldest and
most detailed one by Jacques Cellier (ca. 1550-ca. 1620),
and quoted all the descriptions and the paraphrases of the
inscriptions (fig. 2). Furthermore, he established a suc-
cession for the masters: Jean d'Orbais (in the upper right-
hand corner) began the coiffe, that is, the chevet; Jean le Fig. 1. Reims Cathedral, former labyrinth, drawing by J. Cellier
Loup (in the upper left-hand corner), in office for sixteen (courtesy: Bibliotheque Nationale).
years, began the portals; Gaucher de Reims (in the lower
left-hand corner), in office for eight years, worked on the
portals and the voussures; and Bernard de Soissons (in the about 1290,4 Demaison was able to fix approximate dates
lower right-hand corner), in office for thirty-five years,
for the masters as follows: Jean d'Orbais, 1211-1231; Jean
made five of the nave vaults and the western rose window.
le Loup, 1231-1247; Gaucher de Reims, 1247-1255; and
And finally, since he assumed the labyrinth was made
Bernard de Soissons, 1255-1290. Some confirmation of
these dates was provided in 1898, when Demaison dis-
1. See W. H. Matthews, Mazes and Labyrinths (London, 1922), covered Bernard's name in a tax list of 1287.5
pp. 60-61 and passim.
2. For example, A. P. M. Gilbert, Description historique de l'eglise
metropolitaine de Notre-Dame de Reims (Reims, 1825), pp. 26-27; 4. Demaison, 'Les architectes . . .', p. 23; the assumption was
P. Tarbe, Notre-Dame de Reims, 2d ed. (Reims, 1852), pp. 11-1112; based on the similarity of the Reims labyrinth to the one at Amiens,
L. Paris, 'Notice sur le dedale ou labyrinthe de l'eglise de Reims', which was made in 1288.
Bulletin monumental xxII (1856), 540-551, the first serious consid- 5. Louis Demaison, 'Nouveaux renseignements sur les architectes
eration of the subject; and Ch. Cerf, Histoire et descriptionde Notre- de la cathedrale de Reims au moyen age', Bull. arch. (1898), pp.
Dame de Reims (Reims, 1861), I, 77-80 and 395-396; II, 229-230. lx-lxi and 40-48; Demaison later discovered the same master's name
3. Louis Demaison, 'Les architectes de la cathedrale de Reims', in 1282: see his communication to the Societe nationale des anti-
Bulletin archeologique, Comite des travaux historiques et scientifiques quaires de France, Bulletin (1931), pp. 150-153 and 'Nouveau ren-
(1894), pp. 1-40; see his communication, ibid. (1891), xxxiii-xxxiv. seignement sur Bernard de Soissons, maitre de l'oeuvre de la ca-
He maintained and defended the position stated in 1894 throughout thedrale de Reims', Nouvelle revue de Champagne et de Brie ix
the next forty years. (1931), 186-187.

18

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
 19

JEAN D'ORBAIS
C: ... l'image d'un Jehan d'Orbais maistre des dits ouvrages qui encommencea la coiffe de l'eglise.
L: La derniere (figure) en remontant est de Jean d'Orbais
R: Cette image est une remembrance de maitre Jean d'Orbois qui fut maitre de l'eglise de ceans ...
H: Cette image est en remembrance de maitre Jean d'Orbais qui fut maitre de l'eglise de ceans ...

JEAN LE LOUP
C: ... l'image d'un maistre Jehan de Loup qui fut maistre des ouvrages d'icelle eglise l'espace de seize ans et commencea les portaux d'icelle.
L: La premiere figure ... est de Jean Loup qui a commence le grand portail.
R: ... Jean Loup qui fut maitre de l'eglise de ceans seize ans et en commence ...
H: ... qui fut maitre de l'eglise de ceans seize ans et encommenca ...

GAUCHER DE REIMS
C: ... l'image d'un Gaucher de Reims qui fut maistre des ouvrages 'espace de huict ans qui ouvra aux vossures et portaulx.
L: La deuxieme (figure) est de Gaucher de Reims qui travailla aux vofutes et aux portaux.
R: Gaucher de Reims fut maitre de l'eglise de ceans sept ans et ouvra a vouzure dou ...
H: qui fut maitre de l'eglise de ceans sept ans et ouvra a vosures ... dor ...
BERNARD DE SOISSONS
C: ... l'image d'un Bernard de Soissons qui fit cincq voutes et ouvra a 1'0 maistre de ses ouvrages l'espace de trente cinq ans.
L: La troisieme (figure) est de Bernard de Soissons qui fit cinq voutes et la grande roze du portail.
W: Cette image est en remembrance de maitre Bernard de Soissons maitres de l'eglise de ceans.
R: Cette image est la remembrance de maitre Bernard de Soissons qui fut maitre de l'eglise de ceans ... il fit cinq voutes.
H: Cette image est en remembrance de maitre Bernard de Soissons qui fut maitre de l'eglise de ceans ... fit cinq voutes.

Fig. 2. Table of paraphrases of the labyrinth inscriptions. (c) Cocquault, d. 1645. (L) Lacourt d. 1730. (w) Weyen, first quarter,
eighteenth century. (R) Robin, 1779. (H) Have, 1779.

Since 1894, however, discord has reigned over the nently logical to us to follow the path of the maze-this
proper interpretation of the labyrinth. At first, such men is suggested by the very number of scholars who have
as Saint-Paul, Kunze, and Brehier proposed only minor subscribed to Deneux's hypothesis-but there is in fact
changes in Demaison's hypothesis.6 Then a fundamental no reason whatever to assume that the late thirteenth-
disagreement arose when Henri Deneux suggested, in century designer of the labyrinth also found it so: maze
1920, that the correct order of succession was the one and masons may well have been unrelated except in the
formed by the labyrinth path itself.7 This would place simplest physical sense. At Amiens all the architects were
Bernard de Soissons first or last, and since he was alive in represented in the central cartouche of the labyrinth there
1287, the order should be Gaucher (1211-1219), Jean le together with the founding bishop, rather than individu-
Loup (1219-1235), Jean d'Orbais (1235-1255), and Ber- ally, at special points along the path. Both of the latter
nard de Soissons (1255-1290). The coiffe built by Jean features seem unique at Reims. Moreover, neither the
d'Orbais would then have been the upper stories of the general nor the specific purpose of the labyrinth in the
chevet, according to an interpretation of the word once mediaeval church has yet been satisfactorily explained,9
propounded by Demaison himself.8 It may now seem emi- whereas the inscriptions and images at Reims clearly had
a commemorative function.10 The images may have been
6. A. Saint-Paul,'La cathedralede Reims au XIIIe si&cle',Bull. placed in certain corners or in a certain sequence in re-
mon. LXX (1906), 297-322; H. Kunze, Das Fassadenproblem der sponse to some contemporary evaluation of the architects'
FranzisischenFriih- und Hochgotik(Leipzig, 1912), pp. 44-75. See respective merits, or even, as Saint-Paul suggested,11 from
also W. Noack,'Aufgabenund Problemearchitekturgeschichtlicher a deep-seated liturgical habit that prompted one always
Forschung',FestschriftfiirA. Goldschmidt
(Leipzig,1923), pp. 116- to start any movement in the far end of the church and
125, esp. 122-123; L. Brehier, La cathedrale de Reims, une oeuvre on the epistle side (in the case at hand, with Jean d'Or-
francaise (Paris, 1916), pp. 25-40.
7. Henri Deneux, 'Chronologie des maitres d'oeuvre de la cathe- bais). The labyrinth path therefore seems irrelevant in
drale de Reims', Bull. soc. nat. ant. France (1920), pp. 196-200. determining the order of succession of the thirteenth-
This article was perhaps stimulated by Demaison's communication century masters.
on technical terms to the same revue in 1919 (pp. 233-248), as well
as by Deneux's appointment as chief architect in charge of restoring 9. For a discussion of the question, see Matthews, Mazesand Laby-
the cathedral after the first World War. See Demaison's two an- rinths, pp. 66-70, with bibliography.
swers, ibid. (1920), pp. 200-201, 236-242. lo. The phrase noted by Weyen in the early eighteenth century
8. See Demaison, 'Les architectes . . .' p. 24, note 1; Demaison, (see fig. 2) removes any doubt that Robin and Have might have
Bull. arch. (1919), pp. 237-240; Deneux, 'Chronologie des maitres copied one another.
. . ', pp. 198-199; and Demaison, Bull. arch. (1920), pp. 240-241. 11. Saint-Paul, 'La cathedrale de Reims . . .', pp. 297-298.

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
20

In 1943 Marcel Aubert proposed that the second mas-

ter was Gaucher (1231-1239), who would have been pri-
marily a sculptor, according to the labyrinth inscription,
and who would have carved the portals now in the north
transept.16 But even this did not exhaust the explanations
and combinations. In 1949 Walter Uberwasser revived
Deneux's order,17 and in 1957 Mme. Wyffels-Simoens pro-
posed a half-Deneux, half-Male solution:18 the unknown
central figure was the first architect, in office for a single
year (1210-1211), and he was succeeded by Gaucher
(1211-1219), Jean le Loup (1219-1235), Jean d'Orbais
(1235-1255), and Bernard. And finally in 1958 Elie Lam-
bert suggested that there might have been a gap between
1241, when the canons were installed in the chevet of the
new cathedral, and the undertaking of a new campaign
about 1247, hence: Jean le Loup (1211-1227), Jean d'Or-
bais (1227-1241), Gaucher (ca. 1247-ca. 1255), and Ber-
nard; and that the central figure was Robert de Coucy
Fig. 3. St.-Remi, Reims, paving stones from St.-Nicaise, now re- (d. 1311), the early fourteenth-century master of the
arranged (photo: author). cathedral.19
The order that Demaison proposed in 1894 was based
on a sensible interpretation of the information available
from the labyrinth and it happened to form a regular,
counterclockwise progression. At that time Demaison ob-
Paul Vitry could not decide between the alternatives
viously had no idea that his article would prove to be the
proposed by Demaison and Deneux.12 In 1921, however,
touchstone for so much scholarly disputation, or that the
Emile Male suggested that the first architect was neither
nor Gaucher, but was represented by the
labyrinth would be treated as an alternate and shorter
Jean d'Orbais
route to the normally lengthy business of isolating and
central, unnamed figure of the labyrinth.13 It is hoped that
the present essay will demonstrate how unnecessary this dating the campaigns of construction. This, it seems to
me, is the error into which many of the recent hypotheses
suggestion is. Then in 1927 Erwin Panofsky proposed the have fallen. Because our knowledge of the images and in-
last remaining possibility, Jean le Loup, as the first archi-
tect and justified his choice in the following manner.14 scriptions is very incomplete, the labyrinth cannot serve
as a point of departure for the study of the monument;
Jean le Loup began the portals, according to the laby-
rinth inscription, but the project was altered and the quite the contrary, it must itself be interpreted in terms
of the traditional analysis of the monument and of the
sculpture moved to the north transept, while a new west
facade was started under a new master; since the latter
exerted an influence on sculptors at Mainz and Bamberg 16. Marcel Aubert, 'Les campagnes de construction de la cathe-
in the mid-1230s, Jean le Loup must already have been drale de Reims', Comptes-rendusde l'academie des inscriptions et belles-
out of office for a few years, and his sixteen years of work, lettres (1943), pp. 203-209 and Bulletin de la societedes amis du musee
de Dijon (1946-1948), pp. 24-26; more recently, this author re-
placed at the start, would fit these requirements perfectly turned to Demaison's proposal of 1894 in 'L'ordre des architectes
(1211-1227). Jean d'Orbais would then have been in de la cathedrale de Reims', Bull. soc. nat. ant. France (1955), 165-
charge from 1227 to about 1247/1250 (and coiffe must be 166; 'Les architectes de la cathedrale de Reims', Bull. mon. cxiv
translated in Deneux's manner), Gaucher from 1247/1250 (1956), 123-125; 'La construction au moyen age', Bull. mon. cxix
to 1255/1258, and Bernard from 1255/1258 to 1290/ (1961), 28.
17. Walter Uberwasser, 'Massgerechte Bauplanung der Gotik an
1293. This hypothesis becomes untenable, however, if
Beispielen Villards de Honnecourt', Kunstchronik ii (1949), 200-
Jean d'Orbais is considered the first architect.15 204, esp. 203-204.
18. M.-L. Wyffels-Simoens, 'Note sur le labyrinthe de la cathe-
drale de Reims', Gazettedes beaux-arts, series 6, xLIx (1957), 337-
12. Paul Vitry, La cathedrale de Reims (Paris, 1915-1919), I, 10. 340.
13. Emile Male, 'La cathedrale de Reims (a propos d'un livre re- 19. Elie Lambert, 'Le labyrinthe de la cathedrale de Reims, nou-
cent)', Gazettedes beaux-arts, series 5, LXIII (1) (1921), 73-88, re- vel essai d'interpretation', Gazettedes beaux-arts, series 6, LI (1958),
peated in his Art et artistes du moyenage (Paris, 1927), pp. 231-234. 273-280; this opinion seems to be accepted by M. Eschapasse in the
14. Erwin Panofsky, 'Uber die Reihenfolge der vier Meister von official guidebook of the Commission des Monuments Historiques,
Reims', Jahrbuchfiir Kunstwissenchaft (1927), PP. 55-82. La cathedrale de Reims (Paris, 1961), p. 11. See, however, tlie Lam-
15. Robert Branner, 'Jean d'Orbais and the Cathedral of Reims', bert, 'La construction de la cathedrale de Reims au XIIIe siecle',
Art Bulletin XLIII(1961), 131-133. Gazettedes beaux-arts, series 6, LVII(1961), 217-228.

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
 21

texts. Having elsewhere studied a number of the texts,20

and one of the major changes of plan,21 I shall endeavor f 1- 122.0

here to provide new arguments to support the main lines E s <8

1f3 U17 7 ZO -I 12336 m
of Demaison's hypothesis.
The labyrinth inscriptions have often been analyzed, e ~ ~C??7 - IS4-
00 | e?~ ' 1
1:'::1IZ36
0
but one aspect of their physical state has not yet been
sufficiently emphasized. The maze was thirty-four feet
wide, with eleven-inch paths outlined by 'lines' of bluish-
black stone that were probably about four and one-half
inches wide.22 It was thus a sort of cosmatesque design, O 0Ow ,.
* 0
like the majority of labyrinths from the twelfth and thir-
teenth centuries in northern France. But the inscriptions 0 0
and the figures, rather than being executed in mosaic,
were probably chiseled into the limestone and filled with
lead or mastic. Cocquault, whose description of the work
is undoubtedly the best, described the central image as
'insculpee', which suggests a broad, shallow depression
something like a large enamel cloison, into which a thin
layer of material had been poured.23 This would explain
the solid bodies of Cellier's drawing. The heads, hands,
feet, attributes, and inscriptions, on the other hand, were Fig. 4. Reims Cathedral. (A) campaigns of construction at ground
most likely executed in filled lines. This was a standard level. (B) at triforium and main vault level.
technique, of which both the tombstone of Hugh Liber-
gier and the pavement from St.-Nicaise at Reims are con-
a half other parts deteriorated beyond recognition or van-
temporary local examples (fig. 3).24 Like the 'cloisonne',
it was very vulnerable to the passage of feet, since both the ished, as can be seen from figure 2. The losses very likely
limestone and the filling were soft: Libergier's tombstone included Jean d'Orbais's 'coiffe', Jean le Loup's 'portail',
had been almost completely denuded of its lead by the Gaucher's 'portaulx', and Bernard's 'O' and term of office.
nineteenth century, and the designs are nearly effaced on Gaucher's term was also altered from vIII to vII between
several of the St.-Nicaise paving stones. The labyrinth about 1600 and 1778, a point of capital importance to
suffered in the same way. In Cocquault's time the inscrip- which we shall return below.25
tion surrounding the central figure was so worn that it The documents inform us that the Gothic cathedral was
could no longer be read, and during the next century and undertaken immediately after the fire of 6 May 1210 and
that the first part of the work above ground was begun
exactly a year later.26 About 1220 the upper stories of the
20. RobertBranner,'Quelquesdates pour servira l'histoire de la
chevet were in preparation, and the termination of the
constructionde la cathedralede Reims (1210-1241)', Memoiresde la building was in sight in the 1230s, when the civic dis-
societe'd'agriculture, commerce,sciences et arts du departement de la turbances of November 1233-January 1236 seem to have
Marne LXXV (1960), 78-81, and 'Historical Aspects of the Recon-
interrupted the work. On 7 September 1241 the chapter
struction of Reims Cathedral, 1210-1241', Speculum xxxvI (1961),
was installed in the choir, and from that time on the work
23-37, esp. n. 5.
21. See Robert Branner, 'The North Transept and the first West seems to have progressed at a somewhat slower rate. Quests
Facades of Reims Cathedral', Zeitschrift fur Kunstgeschichte xxiv for funds were made in 1246, in 1251, and in 1295, and by
(1961), 220-241. 1299, when the southwestern tower was undertaken, the
22. Ch. Cerf, Histoire .. ., I, 78, mentioned fragments still visible
body of the nave and facade had been terminated.27
at Reims, but by Demaison's time they seem to have vanished. See
Demaison, 'Les architectes . . .', p. 15, n. i.
23. The term meant sculpted, engraved, or imprinted (Huguet, 25. See Demaison, Bull. arch. (1920), p. 238. These changes make
Dictionnaire de la languefranfaise du XVIe siecle, ad verb. insculper). it likely that the labyrinth had sustained losses before the time of
Martene and Durand described the pavement of Arras Cathedral, Cellier, namely, Gaucher's attribute and the head of the central
which was similar to the Reims labyrinth, as 'en bosse', that is, with figure, and perhaps Jean d'Orbais's tenure had vanished before the
a depressed design. See L. Serbat, Bull. mon. LXXXVIII (1929), 372. time of Cocquault. Demaison also noted that Cocquault's text on
See also J. Gailhabaud, L'architecture du Ve au XVIIe siecle (Paris, the Jean d'Orbais inscription was written at two times, as if the
1858), III, 'Dalles tumulaires' and E. Ame, Les carrelages dy moyen good canon succeeded in deciphering a worn portion only after
dge (Paris, 1859), pp. 57-64, for the technique and for other medi- considerable effort (Demaison, 'Les Architectes . . .', [1894], p.
aeval examples. 43, n. 2).
24. See L. Demaison in Congres archeologique LXXVIII(1911), 1, 26. See Branner, 'Historical Aspects ...', pp. 23-24.
illustration between 48 and 49, and 97-98, and E. Ame, Les carre- 27. See, inter alia, L. Demaison, Album de la cathedrale de Reinms
lages. . ., pp. 70-71. (Reims, 1899), pp. 7-8.

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
22

Fig. 5. Reims Cathedral, capital from first campaign, south aisle of Fig. 6. Reims Cathedral, capital from second campaign, nave, south
chevet (photo: author). file (photo: author).

The analysis of the monument reveals that there were the upper stories to the east, it included the lowest parts
four major campaigns of construction in the thirteenth of two more bays of wall in the nave aisles and twelve of
century (figs. 4, 5).28 The first comprised the ground the piers (fig. 6). The projected nave was probably to have
story of the chevet and the transept, including the periph- comprised eight bays, with a pair of heavy, rectangular
eral wall of the first three bays of the nave aisles (counting piers at the western end, but the foundations for this por-
from the east).29 At that time, only simple doorways were tion do not seem to have been laid.33 Moreover, the work
above ground also did not come to its full, logical conclu-
planned in the transepts, the Porte romane giving access
to the cloister lying to the north and an even less elaborate sion due to the riots of 1233. By that time, preparations
had been made to vault the chevet and the south transept,
opening in the east bay of the south transept connecting
the cathedral with the archbishop's palace. Some sculp- including the construction of the vault springers. The
ture for the future west facade was also begun in this north transept, the crossing, and the six bays of the nave,
campaign.30 however, were not so advanced; the Remois passage in the
Henri Deneux showed conclusively that the triforium side-aisle wall, for instance, had not yet been built in the
and clerestory of the chevet and transept were erected by last two bays of the north aisle, and the aisle vaults there
a new master in another campaign of construction.31 This could not have been erected. It was during this campaign,
must have been inaugurated about 1220.32 In addition to about 1225, that the plan of the west facade was laid out
in detail and the sculpture for the portals was undertaken
in earnest.34 Only two or three years later the project was
28. See Aubert,'Lescampagnes . . ', and H. Deneux, Modifica-
tions apporteesa la cathedralede Reimsau cours de sa construction rejected, the sculpture was fitted into two new portals
du XIIIe siecle au XVe siecle', Bull. mon. cvI (1948), 121-140. opened to receive it in the north transept, and another
29. Aubert, 'Les campagnes. ..', p. 206, is certainly incorrect in project was designed for the west fagade. It must again be
attributingthe three south aisle responds of the nave to another kept in mind, however, that the architecture of the west-
campaign; the eighth and ninth piers from the crossing, in the south ern end of the cathedral was not erected at this time.
file, are also incorrectly attributed, since they are identical in time
with their opposite numbers on the north.
30. The prophets now adorning the right jamb of the right portal
of the present west facade, and the peak of the Last Judgment 33. Panofsky,'Uberdie Reihenfolge. . .', p. 56 and Uberwasser,
tympanum now on the north transept are generally considered to 'MassgerechteBauplanung. .', complicatethe situationneedlessly
be the oldest pieces of thirteenth-century sculpture at the cathedral. by dividing the total numberof eight bays into seven for the nave
31. Deneux, 'Modifications . .., pp. 123-125. and one for the facade. A straightforwardnumberingsystem is
32. Branner, 'Historical Aspects .. .', pp. 30-31; Zeitschrift xxiv used here.
(1961), 230; and 'Paris and the Origins of Rayonnant Gothic Ar- 34. See Teresa G. Frisch, 'The Twelve Choir Statuesof the Ca-
chitecture down to 1240', forthcoming in the Art Bulletin. thedralat Reims', Art Bulletin XLII(1960), 1-24, for the date.

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
 23

Fig. 8. Reims Cathedral, capital from start of fourth campaign, nave,

south aisle, seventh respond from transept (photo: author).

I -
Fig. 7. Reims Cathedral, capital from third campaign, nave, north
side, respond above sixth pier from crossing (photo: author).

After the peace of 1236, the chapter apparently felt it

was imperative to put the monument into service as quick-
ly as possible. The presence of another group of workmen
and another master is indicated by the sculpture of the
capitals in the triforium and clerestory of the first six bays
of the nave (fig. 7).35 Unlike the elaborate capitals made
in both of the preceding campaigns, those of the third
campaign are simple and severe, with long, spiny crockets
but few leaves, if any, on the core. By 1241 this shop, un-
doubtedly working from scaffoldings left in place by its
predecessor, had completed the vaults of the chevet and Fig. 9. Reims Cathedral, capital from end of fourth campaign, nave,
south transept on a new and higher curve, as Deneux north side, respond above northwest tower pier (photo: author).

pointed out;36 it had also vaulted the north transept and

crossing and it had finished the first three bays of the
nave, which were probably closed off by a temporary wall.37 After 1241, it continued work on the next three bays of
the nave and by the early 1250s it seems to have terminated
those parts that had been laid out and started before 1233.
35. The bases executed by the third shop have no cavetti. This The masonry of the facade and of the four western bays
feature generally vanished from bases in northern France in the
decade between 1235 and 1240.
of the nave is uniform and belongs to a single campaign,
36. Deneux, 'Modifications . .', pp. 126-128. the fourth in the series. This is indicated by the design of
37. A thin closing wall would be a practical necessity to protect the window tracery, by the bases (without cavetti but not
the occupied parts of the church from weather. But such a wall did
resembling the two-torus type of the third campaign) and
not have to rest on heavy foundations, such as Deneux found lying
across the west side of the first nave bay (H. Deneux, Dix ans de particularly by the capitals, in which the crockets have
almost entirely given way to groups of very naturalistic
fouilles d la cathedrale de Reims, 1919-1930 [Reims, 1944], folding
plan); the latter must have been intended as a transverse link be- leaves (figs. 8, 9).38 At first glance the latter seem quite
tween the longitudinal foundations beneath the piers and aisle walls.
Although there are no longer any physical indications of the jube
on the piers between the third and fourth bays of the nave, that is this line seem always to have rested on corbels, to permit the choir
where it was located (see the Lepautre engraving of the coronation stalls to be placed farther apart from one another and to enlarge the
of Louis XIV in M. Sartor, Les tapisseries, toiles peintes et broderiesde area of the choir proper.
Reims [Reims, 1912], fig. 8) and where the liturgical choir at Reims 38. A longer list of changes will be found in Deneux, 'Modifica-
always terminated. The colonnettes on the four piers to the east of tions . .', pp. 128-132.

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
24
close in style to the capitals of the Sainte-Chapelle in and his term of office would have extended from 1220 to
Paris .(1240-1248) . But the technique of cutting, the 1236,just after the end of the civic disturbances at Reims.43
epannelage or distribution of ornament over the core, and Since Bernard de Soissons was alive in 1287 and hence
the 'feel' of the leaves differ at Reims and must be com- could not have assumed office before 1252, the third ar-
pared with such works as the triforium in the chevet of chitect of the cathedral must have been Gaucher de Reims,
Amiens cathedral (ca. 1250), St.-Urbain at Troyes (1262 whose eight years of work would span the years 1236-
and after), and the ambulatory of St.-Germain at Auxerre 1244. From 1236 to 1241 he finished the chevet, the tran-
(1277 and after).40 This suggests that the fourth campaign sept, and the first three bays of the nave, and according
at Reims was inaugurated only in the 1250S and extended to the labyrinth he was primarily occupied by the sculp-
over the next several decades. Such a date, however, ture for the west facade. In this light, three years would
largely concerns the architectural framework (including scarcely have been sufficient to complete the remainder
the lower groups of reliefs on the verso of the facade), of the nave. There is, however, some question as to the
since many of the portal statues had been carved in the accuracy of the labyrinth inscription, or rather of the
1230s and 1240s.41 paraphrase of the inscription. As was noted above, the
The major campaigns at Reims can therefore be dated portion that mentioned Gaucher's tenure was well worn;
with relative clarity as follows: I, from 1210 to about 1220; the viII, for instance, had already degenerated to vii be-
II, from about 1220 to 1236, with little or no work being tween about 1600 and 1778. It is not impossible that be-
done during the riots; III, from 1236 to the mid-1250s; and fore 1600 an x had been lost and that the number was
iv. from about 1255 to the 128os or 1290s. It is not diffi- originally xvIII rather than vIII.44 Eighteen years would,
cult to see a correspondence between these dates and the in fact, span the period from 1236 to 1254 and give suffi-
terms of office mentioned in the labyrinth. cient time for Gaucher to have completed those parts of
Jean d'Orbais was probably the first master of the Goth- the nave that had already been started, as well as to have
ic cathedral, since the pattern he was shown drawing in continued the portal sculpture which the labyrinth in-
the labyrinth seems to have represented the geometric scription assigned to him.
procedure used in the layout of the hemicycle of the mon- 1254 is a perfectly acceptable date for the accession of
ument.42 He would therefore have designed the ground Bernard de Soissons, who would have died or left the
plan and built the lower portions of the chevet and tran- chantier in 1289, two years after his name is last recorded
sept, and 'coiffe' would mean nothing more or less than in the documents. It was therefore he who undertook to
'chevet' in a general sense. Since no length of office is construct the western bays of the nave and the architec-
known for him, it is logical to assume that he was replaced ture of the present facade. The span of time between the
by a second master about 1220, when the second cam- designing and the execution of this project (before 1233-
paign was undertaken. after 1254) was not unusual, for even longer gaps are
During the second campaign two west facades were suc- known at Paris and at Laon in the late twelfth century.45
cessively designed and the sculpture for them was under- There seem also to have been changes of plan at Reims
taken. The first was the false start of about 1225, the re- during the interval, as there were at both Paris and Laon,
mains of which can now be seen in the north transept, and and of course Bernard de Soissons himself altered the
the second was the design of about 1228/1230, which work as it progressed.46
ultimately formed the basis of what we see today at the
western end of the cathedral. But architecturally both
43. It would not havebeen unusualfor the mastermason to have
projects were probably quite similar to one another and remainedon duty during the interruptionof work caused by the
may well have been due to the same hand. This would riots. A MasterArnoul the Stonecutter,who lived on the parvisand
have been Jean le Loup's, 'qui ... commencea les portaux', was very likely engagedin workon the cathedral,raised?2o Pruv.
on his house in November 1234, see Demaison, Cong. arch. (1911),
p. 159. This suggests both that work was in fact suspended and that
even some of the skilled labor stayed on at Reims.
39. D. Jalabert,'La flore sculptee de la Sainte-Chapelle',Bull. 44. This number of years was also given to Gaucher by several
arch. (1935), pp. 739-747. authors in the early nineteenth century, such as Gerusez, Povillon-
40. For Amiens, see G. Durand, Monographiede l'eglise Jotre- Pierard, and Tarbe (see Demaison, 'Les architectes .. .', p. 18, and
Dame, cathedraled'Amiens(Paris-Amiens,I, 1900), figs. 86 and 96; Demaison, Bull. arch. [1920], pp. 238-239), but it was obviously
for Troyes, F. Salet in Cong.arch. cxiiI (1955), fig. p. 108;for Aux- a simple error for eight and has nothing to do with the present
erre, Branner,Burgundian GothicArchitecture(London, 1960), pl. argument.
33; resemblancesalso exist with the second campaignat Strasbourg, 45. For Pa is, see M. Aubert, Notre-Dame de Paris. Sa place dans
see Ch. Wittmer,La cathedralede Strasbourg(Lyon, 1950), fig. p. 31. l'histoire de l'architecture, 2d ed. (Paris, 1929), p. 39; for Laon, see E.
41. Aubert,'Les campagnes.. .', p. 208, has suggested that Jean Lambert, 'Les portails sculptes de la cathedrale de Laon', Gazette
le Loup implantedthe foundationsfor the present facade between des beaux-arts, series 6, xvI (1937), 83-98.
1239 and 1255, but there is no evidencefor this in Deneux, Dix ans 46. For a major alteration to the buttresses, see Doris Schmidt,
defouilles .... 'Portalstudien zur Reimser Kathedrale', Miinchner Jahrbuch der
42. Branner, 'Jean d'Orbais . . .' bildenden Kunst xi (1960), 14-58.

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions
 25
It thus seems unnecessary to revise Demaison's pro- who was in office when the fire broke out in 1210, or
posed order of succession for the architects, or to dwell Pierre Barbette (1273-1298), or Robert de Courtenay
unduly on the mysterious central figure of the labyrinth. 1299-1324), one of whom was archbishop when the laby-
If the latter represented an architect at all, then it must rinth was made, the same conclusion can be drawn. In
have been one who was in office after 1289 and who was sum, the lost labyrinth of Reims cathedral seems capable
therefore not responsible for any part of the body of the of a coherent, logical explanation, but one that must be
cathedral (Lambert's thesis). If, on the other hand, it was considered a conclusion and not a point of departure for
an archbishop,47 whether Aubri de Humbert (1207-1218), the study of the cathedral.

47. Demaison and C. Enlartboth thought that the dress of the figure. The Cellier manuscriptwas recently reboundand partially
centralfigurewas that of a prelate. See Demaison, 'Les architectes laminated,and a smallpiece of paperwas removedthat once covered
. .', p. 23, and Enlart (1920), p. 242. In fact the only indications the areawhere the head should be in the drawing;but no head was
of dress are three rectangular panels below the neck, which can found beneath. A similar attempt seems to have been made some
indeed be likened to the amict, a strip of the chasuble, and one time ago, when the area was treatedwith chemicals to darkenany
branch of a Y-shaped pallium. See J. Demay, Le costume au moyen lost lines; unfortunatelynone appeared.We must thereforereturn
dge d'apres les sceaux (Paris, 1880), figs. 330, 332, 339, etc., where to the obvious conclusion that Cellier did not draw in the head
the line drawings seem to approximate the forms of the labyrinth becausenone existed in the labyrinthat the time he madehis copy.

This content downloaded from 161.116.100.92 on Wed, 4 Sep 2013 06:50:24 AM

All use subject to JSTOR Terms and Conditions