Construction and Conception Techniques of Residential Buildings and

Urbanism in Medieval Europe around A.D. 1100: The Example of Cluny

(France)
Bernhard Flüge
Everybody knows that the Burgundian abbey of Cluny was one of the intellectual and spiri-
tual centres of Europe during the High Middle Ages. But also the surrounding little town is of
scientific interest. Its earliest, partially preserved upright standing habitable stone buildings
were discovered and documented by the author in the last two decades. The results of this
study help to explain the formation of the European town house and town texture in France,
Southern Germany, and Italy—even Rome, in the mediaeval period. The specific construction
and conception of these houses leads to new conclusions concerning building practices, ex-
tending to urbanism and space mastering, apparently based on Roman construction practic-
es and conception instructions.1

The “House with the round arch gate” from 1090/91 (d)

The “house with the round arch gate”, from 1091 according to dendrochronology, is the old-
est precisely dated stone townhouse in France (Fig. 1).2 This house has been integrated in a
texture of mediaeval and post-mediaeval serial houses characterizing Clunyʼs old town since
the later 12th century. The ground plan of the original house appears in rectangular shape in
the back part of the present building area. The building differs with two marked typological
properties from the well-known Romanesque houses in Cluny, which were all definitively
constructed after 1150.3 First, it was originally a freestanding rectangular building; its recon-
struction requires a completion by exterior stairs. Second, it was separated from the street
by a frontcourt, documented by an archaeological survey in 1997. The existence of an upper
floor is an important typological feature that characterizes all town houses in Cluny as “do-
mus solaratae”.4
The construction of the house consists of double shell masonry. The stones corre-
spond to the local material used for the construction of Cluny Abbey, but are of smaller size,
and were thus certainly less expensive. The arch keys of the gate are comparable to the East-
ern parts of the abbey church Cluny III and may attest the accepted dating in the year 1095
for the choir. Concerning the petit appareil masonry, the small stones were shaped approxi-
mately in a block form by a big hammer, a common technique, dating back to Roman times,

1 Cf. detailed analysis in Flüge forthcoming. The present article abstracts some of the main research results
and focuses them to a new view of mediaeval building and planning practices, that seem essentially based on
late antique knowledge.
2 Archéolabs réf. ARC 99/R2138D, ARC 99/R2138D/2, ARC 02/R2767D/1.
3 Cf. Flüge 2011, passim. The article outlines a novel typology of mediaeval residential buildings (houses
and palaces) in Cluny as examples of a European development (for detailed analysis cf. Flüge forthcoming, esp.
Kapitel V, “Domus solaratae der Periode Cluny III – Typologie” and appendix, documentation pl. 1–50).
4 Concerning the introduction of this historically sourced term, cf. Hubert 1990, passim, in particular 170–
179 and 232.
Fig. 1: Cluny, house 20, rue du Merle.
A. (Previous page) Plan with building evolution. Documentation Pl. 7. Drawing: Bernhard Flüge.
B. (Above left) Façade of the initial building (Kernbau) from 1090/91 (d) after readaptation of the surround-
ing level. Photo: Hans-Peter Vieser.
C. (Above right) Initial building from 1090/91 (d). Isometric reconstruction, view from SW. Documentation
Pl. 22. Drawing: Bernhard Flüge.

Fig. 2: Comparison of masonry techniques (clock-

wise from top-left).
A. Cluny, house 20, rue du Merle. Initial building
from 1090/91 (d). External West wall. Characteris-
tic petit-appareil masonry in long courses of oolith
stones with insular micrit elements. Micrit is absent
in the Cluny houses of the 12th century and later.
Photo: Bernhard Flüge.
B. Cluny, house 20, rue du Merle. Initial build-
ing from 1090/91 (d), front side. Masonry and sin-
gle-course foundations. Photo: Bernhard Flüge.
C. Augusta Raurica, external wall of the theatre. Pe-
tit-appareil masonry (stones shaped as blocks by a
hammer) with pietra-rasa plaster and stilatura. Pho-
to: Bernhard Flüge.

particularly in the provinces of the Roman Empire. Perhaps the use of the hammer, called la
mace in old French, derived from the popular Latin word mat(t)ea, was so characteristic for
the craftsmen, that an etymological relation between mace, “hammer”, and maçon, “mason”,
Fig. 3: Cluny, house 20, rue du Merle
A. Conceptional reconstruction of the initial building
from 1090/91 (d). Documentation Pl. 23. Drawing: Bernhard Flüge. B. Documentation Pl. 23, detail. Drawing:
Bernhard Flüge

may be established: Both terms have been transmitted since the 12th century.5 The masonry
is arranged in long courses of fluent and slightly moving appearance (Fig. 2). It is bonded by
an earth or loam mortar, which actually characterizes the period, and contains only few addi-
tions of lime and sand. The foundations are remarkable, because foundation trenches do not
exist. This phenomenon could trace back to the older local building tradition, with wooden
sill-and-post constructions.6 The foundation of the house consists of one only brick-course
just put on the ground, whilst the houses built in Cluny from the 12th century forth dispose
always of foundation trenches. Another difference is the appearance of wood within the con-
nection of the house from 1091, so the lintel of the gate and a wall plate serving as support
for the ceiling beams. Ashlars are exclusively used for openings: the radially dressed arch
stones were certainly fabricated in the masons’ workshop of the abbey. Some vestiges of wall
plaster have been preserved in the interior of the house. It is a pietra rasa plaster, executed
after an originally Roman technique. This plaster can be imagined similar for the interior and
exterior building surfaces. With a depth of 80 cm, the front of the house is the thickest wall. It
contains the biggest stones, and the round arch’s depth almost matches the dimension of the
sidewalls. The position of the gate is absolutely symmetrical. The house front contributes a
monumental feature to the small building. All these details express a representative ambition
of the architecture, attained with a maximal cost effectiveness.
The rectangle of the ground plan is 6.27 m on 10.80 m large and was defined by two
diagonals of similar length (Fig. 3). The diagonals contain approximately the double breadth
of the front, with a contraction difference of 8 inches. The declivity of the natural terrain ex-

5 Cf. Dauzat 1971, 434 (with a different etymological proposition for maçon) and 450 besides <http://ielex.
mpi.nl/cognate/2068/>, 08/17/2012, (mat(t)ea).
6 For vestiges of wooden constructions, cf. Roiné 1988/1993.
plains this difference: The house was not laid out after a horizontal projection like in mod-
ern times, but directly on the 11 degree inclined surface of the natural ground, according
to the topography. The horizontal projection length of the plan appears as the shorter side
adjacent to the inclined hypotenuse line
on the terrain surface, if both lines are
imagined to form an angle in the longi-
tudinal section of the building. The half
breadth of the house from 1091, 3.14 m,
can be interpreted as 1 perch, equivalent
to about 10 feet or 3.4 yards. These ba-
sic units correspond to the dimensions
of several independent architectural
elements of the house, such as niches.
Only the arch of the gate seems to have
been dimensioned with a smaller mea-
Fig. 4: Cluny III, Aula from 1107/08 (d). Eastern eaves side
sure-of-foot, around 30.4 cm instead of
(front of the building) and lateral South wall, tympanum
with lion sculpture. Photo and layout: Bernhard Flüge. around 31.5 cm. Like the longer foot,
the smaller unit appears also elsewhere
in Burgundy. The presence of two slightly differing dimension systems, corresponding to
the contributions by different guilds to one same construction, is one of several particular
examples for the flexibility of building practices around A.D. 1100.
 Fig. 5: Cluny, Abbey. Axial correlation
between aula III and abbey church II.
Documentation Pl. 1, detail. Drawing:
Bernhard Flüge.

The “Aula imperialis” of the

Cluny III period from 1107/08
(d)

Twenty years after the beginning

of the construction site of Clu-
ny III, around 1088, the second
tallest building of the mediaeval
abbey was erected: the so-called
“Écuries” or “Hôtellerie de Saint
Hugues” (Fig. 4). Scholars have
considered this building as a
guest house. But it is not a guest
house type, nor does the Cluniac guest house seem to have been situated at the Écuriesʼ
place.7
The rectangular Écuries hall building is composed of a basement and a main level, 53
m long, nearly 15 m wide without annexes, and nearly 16 m high under the roof. Windows
are arranged on three sides, the long east front and the lateral north and south gable walls.
Together with the central gate of the long front and the pillar axes in the interior of the base-
ment, the windows distribution defines a transversal orientation of the hall building. The
presence of a big hall with basement, the transversal orientation of the building, and the for-
mer presence of annexes, e.g. with stairs and a big loggia, correspond to a palace, and more
precisely, to an aula type.8 The transverse axis of this aula adopts the longitudinal axis of the
Cluny II abbey church, situated on the opposite side of the main abbey court (Fig. 5). The
mediaeval author Gilo writes about an Aula imperialis, which the Abbot Hugues de Semur
had begun shortly before the end of his lifetime (he died in 1109). This information does not

7 For the typological definition of the Écuries, see Flüge 2011, 300–303. For typology and location of aula
and guest house, the 12th c. Canterbury plan (“Eadwine Psalter” (around A.D. 1155–70), Cambridge, Trinity
College, Ms R.17.1 fol. 285) may be compared to the buildings and sources of Cluny. This discussion is part of
the author’s thesis (Flüge forthcoming).
8 Cf. Günther Binding in Lexikon des Mittelalters I (1999), col. 1234: The term “aula” designates “[...] seit
merow[ingischer] Zeit zumeist einen profanen Saalbau bzw. die Abtswohnung (St. Galler Klosterplan), jew-
eils gleichbedeutend mit palatium, seit karol[ingischer] Zeit vornehmlich profane Anlagen mit einem Saal im
Obergeschoss, in Pfalzen die aula regia mit dem Saal des Reiches (Aachen, Ingelheim, Frankfurt, Goslar). Als
Bauform ist die karol[ingische] aula regia ein querrechteckiger, zumeist ost-west-gestreckter Bau mit Unter-
geschoß und einem über ein solarium (wohl teilweise hölzerner über Außentreppen erreichbarer Vorbau)
zugängl[ichen] Obergeschoß, dessen Decke oder offener Dachstuhl von einer Mittelstützenreihe getragen wird.
[...] Der quergerichtete Saal wird auch in otton[isch]-sal[ischen] (Paderborn, Bamberg) [...] beibehalten und in
den Palas auf Burgen übernommen [...].”
 Fig 6: Lion acroteria
A. Cluny III, Aula from 1107/08 (d). South gable wall with
corner lesene, console cornice, and tympanum lion relief. Re-
markable: the slight prominence of the tympanum surface in
relation to the clerestory. Photo: Bernhard Flüge.
B. The 12th c. “AULA NOVA” with lion and dragon acroteri-
ons in Canterbury. Representation of the cathedral domain in
the “Eadwine Psalter” (around 1155–70, Cambridge, Trinity
College, Ms R.17.1, fol. 285). Historical graphic rendering with
unprecise legend reproduction (after Benevolo 1983, 186, Fig.
515, without signature; colour layout: Bernhard Flüge).

mean, in my opinion, that the Écuries was, as is claimed, a metaphoric “imperial habitation”,9
but rather was a new aula, a palatium-type building, which has dated by dendrochronology
to A.D. 1107/08.10

9 For this interpretation of the Latin terms aula imperialis (“kaiserliche Wohnung”), cf. Wollasch 1996, 171.
10 Dendrochronological dating by Christian Dormoy, by request of the author. References: Archéolabs réf. ARC
06/R3275D and Archéolabs réf. ARC 06/3275D/2.
Fig. 7: Cluny III, Aula from 1107/08 (d). Ground plan, frontal view with roof structure and geometrical square
grid for proportion and dimensioning of the building. Documentation Pl. 40 (detail). Drawing: Bernhard Flüge.

Its efficient realization within the construction site of Cluny III might be one of the
reasons that the aula has not been identified during decades of archaeological interventions:
The building features few sculptures, only a three-sided console cornice under the roof, and
the relief of a lion in the southern tympanum. A comparable lion sculpture existed in the aula
of Canterbury and is represented in the 12th century Canterbury plan as an acroterion on
the roof of the “AULA NOVA” (Fig. 6).11 In the Cluniac aula, ashlars are present at the arches
of windows and gates or doors, also at the exterior corners, but not consequently. Several
openings show inexactly fitting arch stones that could be reused. The petit appareil masonry
is partially passed over by a hatchet or a large chisel and was originally covered by a pietra
rasa plaster. This treatment included the big, round basement pillars. The small stones were
easily to process and could be employed anywhere; this might have contributed to a quick
construction process. Another reason for efficiency could have been a concise project, easy to

11 Historic copy of the so called “Eadwine Psalter,” Cambridge, Trinity College, Ms R 171 fol. 285 (after Benev-
olo 1983, 186, fig. 515, without signature; detail. Subsequent layout by the author.
Fig. 8: The “Double-hall-and-high” house (clockwise
from top-left)
A. Cluny, houses 13, place Notre-Dame / 3, rue de la
Barre. “Double-hall-and-high-house” from 1135/36 
(d). Isometric reconstruction, view from North. Doc-
umentation Pl. 39, detail. Drawing: Bernhard Flüge
B. Cluny, houses 13, place Notre-Dame / 3, rue de la
Barre, “Double-hall-and-high-house” from 1135/36
(d) Mediaeval West wall along the ascent of the Rue
de la Barre: Frontal stairs forebuilding with biforium,
central double hall volume, and four level high house
in the back. Photo: Bernhard Flüge
C. Cluny, Place Notre-Dame, West porch of No-
tre-Dame (left), the square fountain in its Louis XVI
state, and houses 11–13, place Notre-Dame (frontal
view). Photo: Bernhard Flüge

mediate and to control. The identical dendrochronological dating results from the basement
and from the roof indicate that the lumber was cut and prepared according to a preliminary
calculation, and successively set in and adjusted during the erection of the building.12 This

12 Concerning this observation, the aula is not a case sui generis in Cluny. The house 23, rue Filaterie / 1, pe-
tite rue des Ravattes, which was equally documented by the author, was erected between A.D. 1193 and around
1205 (d); see Archéolabs N/réf. ARC05/R3325D. The timber of the basement ceiling was cut in the same year
1193 as the timber of the roof structure, including the center purlin and rafters, meanwhile only one of the
roof’s trusses was composed of timber cut after 1200 during the erection of the roof and covered by “older”
timber.
Fig. 9: Cluny, houses 13, place Notre-Dame / 3, rue de la
Barre, “Double-hall-and-high-house” from 1135/36 (d).
A. Basement level, arcades separating the hall volume
from the high house. The deep arcades continue on the
other side of the central wall (left) and support a system
of mural stairs. Photo: Hans-Peter Vieser.
B. Mural staircase between the 2nd and 3rd levels (hall
volume on the left, high house on the right handside).
Photo: Hans-Peter Vieser.

Fig. 10: Cluny, houses 13, place Notre-Dame / 3, rue de la Barre, “Double-hall-and-high-house” from 1135/36 (d).
4th level of the high house, North wall. A. Biforium, column with added base. Originally the windows disposed
of wooden shutters instead of frames, that were fixed on the central column. Photo: Bernhard Flüge. B. Biforium
(exterior), column base. Photo: Bernhard Flüge.
Fig. 11: Cluny, houses 13, place Notre-Dame / 3, rue de la Barre, “Double-hall-and-high-house” from 1135/36
(d). Longitudinal section with foundations realized in two steps (yellow; in dark yellow the excavation coat
in single shell masonry) and roof structure, similar to Roman constructions. Documentation Pl. 30. Drawing:
Bernhard Flüge.

circumstance implicates that the dating by dendrochronology of 1107/08 (d) indicates the
beginning of the construction, not its completion, a fact that is of primary significance for
monumental-archaeological interpretation. According to written evidence, the Vita Hugonis
by Gilo, the abbot Hugo had the courage to begin the aula few before the end of his life—he
died in 1109.13
There are also indicators for a preliminary dimension conception (Fig. 7). The trans-
versal axis is shaped as it were the cardo of the building. This axis passes through the central
gate14 and cuts the round pillar in the centre of the ground plan. The pillar line corresponds
to the longitudinal axis of the rectangular building. The centre-to-centre distance between
the pillars is very constant around 6.32 m, which corresponds to the breadth of the house
of 1091. The geometrical proportion of the plan of 4:1 corresponds to a structural rhythm
of 8:2 and an arithmetical rhythm—in perches—of 16:4. The perch appears equally in the
elevation: two times between the top edge of the foundations to the top edge of the hall floor,
three times between the hall floor and the top edge of the console cornice. This basic dimen-

13 Gilo, Vita sancti Hugonis abbatis II, I, 90, “aulam imperialem” (Wollasch 1996, n. 261).
14 First discovered and documented by the author in 2006.
Fig 12: Cluny, houses 13, place Notre-Dame / 3, rue de la Barre, “Double-hall-and-high-house” from 1135/36
(d). Geometrical analysis of the ground plan. Red: Square grid of a highly probable conception delineation based
on the unit of the double perch (6.30 m), analogue to method and length unit of the conception of the Aula from
1107/08 (d). Shaded: Discrepancies, consciously produced while pegging out the plan on the real terrain, due
to the individual shape of the site. The longitudinal centerline of the grid is thus transformed into an exact angle
bisector. Documentation Pl. 26. Drawing: Bernhard Flüge.
sioning of the building could have been fixed by a small linear rectangular grid design, easy
to mediate and to transfer to the construction site.
The dimension of the walls seems to be defined by a frame contouring the grid, the
exterior line of the frame corresponding to the contour of the building. Proceeding from this
line, the basement walls were constructed 2 feet bigger than the frame width. The frame
corresponds to the wall dimension of the main level. The basic dimensioning appears again
within the total length of the roof construction and in the number of eight sills at the base
of the roof. But the carpenters divided the total length of this specific truss-free construc-
tion into round 50 rafters, respectively 25 beam intervals—a parallel to the predominance of
round distances within the building instructions of the Liber tramitis aevi Odilonis abbatis, a
Cluniac source from the mid of the 11th century.15 The technical solution of the end intervals
in the carpentry, and the choice of big lumber sections, produced execution dimensions in
multiple dependencies, that were adapted with a tolerance of about 2 inches into the pre-
defined total length.16

The “Double-hall-and-high-house” from 1135/36 (d)

In the church square of Cluny stands a building from A.D. 1136 called by the author “Dou-
ble-hall-and-high-house”. This is the second oldest dated town house in France, and also the
largest of all so called Romanesque town houses in Cluny (Fig. 8). The exceptional type com-
posed of two volumes, a lower hall and a higher dwelling house, originates from high nobility
architecture and was in this example transferred into the growing mediaeval town. It is a
house with a double hall on a basement, arranged left and right of a central wall. The frontal
annex is covered in continuity with the hall roof. This connection of a hall and a frontal stairs
annex on a common basement served as immediate prototype for the earliest of the well-
known serial houses after 1150.
Between the hall structure and the 4-level
dwelling house in the back, a sort of bridge construc-
tion on 4 arcades is inserted. In its interior, the circula-
tion between the halls and the high house is organized
in a minimized tube-like space, including the stairs to
the third level of the high house (Fig. 9). A gutter cov-
ers this architectural part, draining the rainwater of the
hall and the high house roofs to the lateral street. The
gutter is composed of stones, the section of which is
comparable to a Roman channel or gutter stone in the
Fig. 13: Villard de Honnecourt, Livre de portraiture (around 1220–
1240). Linear square grid, characteristical of mediaeval building
conception: “vesci une glize desquarie ki fu / esgardee a faire
en lordene dcistiaus” – “See here an orthogonal[ly subdivided]
church, that was envisaged to being made [built] in the order of
Cîteaux”. The central and other lines of the preparatory grid design
15 Cf.partially
were Dinter 1980,
erased203–206.
(scrape marks). B.n.F., Ms fr. 19093 fol. 14v.
16 For detailsand
Transcription cf. Flüge
translation
forthcoming,
by the author.
155–159.
 Fig. 14: A. Gilo, Vita Hugonis, Foun-
dation of the Abbey Cluny III. Text
and Illuminations of this page con-
tain informations of high relevance
for mediaeval conception history,
based on distance units in an or-
thogonal alignment. The protago-
nist Gunzo shows in the right image
the foot as the basic length unit of
the conception of Cluny III. This act
transforms the geometrical square
cord grid of the left image into an
arithmetical scale system. B.n.F., Ms
lat. 17716 fol. 43r.
B. Filarete, Libro di architettura (ca.
1460), rule of thumb “La misura del
pie.” Firenze, BCNF, Fondo Nazio-
nale ms II.I.140 (Codex Magliabe-
chianus) fol. 4r. After Spencer 1965
(facsimile).

museum of Souvigny. A second parallel consists in the fact that both gutters were laid in brick
meal mortar or opus signinum.
The main structure of the house from 1136 consists again of petit appareil mason-
ry, covered by a pietra rasa plaster; ashlars appear in angles, arcades, doors and windows.
Only the biforia columns feature sculptured ornaments; four examples have been preserved.
These columns seem to be prefabricated. During the construction, at least one of the mono-
lithic columns, all of them of approximately equal length originally, was retouched and ex-
tended by addition of a base element to obtain a higher window proportion (Fig. 10).
Several elements of a simple standing roof construction covering the hall section have
been preserved; these permitted the positive reconstruction of five exactly parallel stand-
ing props orientated after the direction of the central wall, and the roof’s conception seems
to have been strictly aligned with the floor plan, the parallel transversal walls of the house
defining a constant width of all poops. Concerning the foundations, the house from 1136,
in contrast to the house from 1091, already disposes of foundation trenches. But in lieu of
a preliminary slope excavation for the whole construction, the common practice in modern
constructions of comparable size, the excavations were realized in at least two steps. The
foundation trench of the exterior wall does not reach down to the final basement level; that
was obtained by digging up to three feet deeper in a second act (Fig. 11). The inner struc-
tures, like the arcade pillars and the central wall, were only founded on the final level, which
retained a rest of the natural slope.
The parallel transversal walls divide the building length from front to back into four
exact quarters, the first of which corresponding to the frontal annex, the two central ones
to the halls volume, and the one on the back side to the high house. Every quarter corre-
sponds to exactly two perches of 6.30 m length traced along the lateral street. The conception
breadth of the building is contained in the garden façade of the high house and measures
exactly four perches (12.60 m).
The trapezoidal enlargement of the building plan towards the front follows the topo-
graphic condition given by the lateral street and modifies a basic conception that consisted of
a square grid of 4:2 double perches (Fig. 12). The middle axis of this square grid was realized
as the angle bisector of the ground plan contour. The central wall was positioned exactly of
this angle bisector, then the “bridge piles” exactly at the half distance between central and
exterior walls. As a consequence, the western bridge pile (marked in green in fig. 12), posi-
tioned at its base on the angle bisector between the central and exterior walls, is orientated
by a slight torsion of its rising structure in the direction of the transversal arcade. This kind
of flexibility appears in many building details, where the orthogonal basic pattern is broken
on account of the trapezoidal realization, particularly in door– and stairways. This building
practice, based on distance tracing, avoids producing waste spaces and at the same time the
basic conception remains perceptible.

Conception and building practices in residential architecture around A.D. 1100…

After this close look at three examples of still upright standing monuments, conclusions
about building practices in residential buildings around A.D. 1100 in Cluny can be summa-
rized as follows:
– The constructions are realized in a relatively simple but specific and flexible way.
The omnipresent petit appareil masonry with pietra rasa plaster and the standing roof struc-
tures take up a building tradition going back to Roman times. The free standing two—or
multiple—level domus solarata type buildings from 1091, 1108 and 1136 correspond to
early medieval upper class house and palace models and are step-by-step transformed into
an urban serial house type during the 12th century. The realized buildings have many eco-
nomically and pragmatically motivated features. Corresponding dendrochronologic dating
results in basements and roofs indicate that the organization of the construction site follows
a preliminary basic dimensioning and calculation.
– Basic dimensioning and building realization correspond to two discernible steps
within the building process. Particularly for larger projects, the dimensions of ground plan
and elevation could have been fixed in a small, linear, square grid design (cf. Fig. 13) and then
transferred to the site. The geometry is rather simple, based on symmetry, the idea of the
right angle, and distance measures. Structural dimensions were defined on the construction
site, as well as the definitive position of walls. Discrepancies to the basic geometry are more
motivated by topographic and pragmatic conditions than caused by error, and appear as par-
allelogram or trapezoid deformations of the rectangle. The arithmetic dimensioning system
adopts an antique tradition, with perch, foot and yard in decimal or respectively duodecimal
relation. There are no signs of a sophistic perfect number conception based on an absolute
“Roman foot” of 29.5 cm, such as supposed for the Cluny III abbey buildings by several au-
thors.17
The only mediaeval document providing information about building practices in Clu-
ny around 1100 is the Latin manuscript 17 716 of the French National Library, in particular
folium 43r with two remarkable illuminations. This source has been considered as a mon-
ument of Cluniac historiography (Fig. 14).18
Meanwhile, its technical aspects have not
been comprehensively explained to mediae-
val studies so far. The possibility of a building
historical interpretation has even been simply
excluded, due to a missing approach.19 After
the preceding description of new-known ex-
emplary buildings of the period and the place,
a new interpretation of the technical aspects
of the document is possible and even essen-
tial. The first illumination on the manuscript
page fol. 43r illustrates a vision of an ancient
abbot of Baume-les-Messieurs, the mother ab-
bey of Cluny, named Gunzo, who lived his old
age in Cluny. The vision instructs him about
the construction of the new abbey church, the
abbey church Cluny III. According to the text
information, the cords stretched by the saints
Peter, Paul and Stephen represent the mea-
Fig. 15: Corpus agrimensorum. Geometric ideal (above)
vs. a consciously “deformed” realization of a colonia, by
respect of topographic conditions. B.A.V., Cod. Pal. Lat.
1564, c. 89v, 1st quarter 9th c. A.D. (after Guidoni 1981,
17 135,
p. Cf. fig.
Conant
189).1968, Appendice B, 140–146, closely followed by Naredi-Rainer 1982, 75–78; cf. also the hy-
potheses of Alain Guerreau, particularly Guerreau 1998, 179–189.
18 Méhu 2002, 56.
19 Cf. Carty 1988, 113.
 Fig. 16: Freiburg (Germany) (counter-clockwise
from top-left).
A.Freiburg (Germany), proposition for the high me-
diaeval town conception around 1100, based on the
coeval geometric systems and working methods dis-
covered in Cluny. Documentation Pl. 6, detail. Draw-
ing: Bernhard Flüge.
B. In Freiburg, the road and water run network was
oriented after the drain direction of the terrain (yel-
low arrow). There are already late Antique centuria-
tion examples showing this practice. Documentation
Pl. 6, detail. Drawing: Bernhard Flüge.
C. Freiburg (Germany), water run (“Bächle”), most
probably a structural element of the town foundation
around 1100 A.D. Photo: Bernhard Flüge.

surement of the quantitas of length and breadth of the building—“longitudinis atque lati-
tudinis metiri quantitatem”. The total dimension of the building is included within the two
terms quantitas and metiri, as well as the aspect of the additive arrangement of the length
unities. These distances define by their rectangular coordination as longitudo atque latitudo
the grid of the architectural project; meanwhile the delimited surface itself is not mentioned.
The painting translates the information of the document’s text immediately into the icono-
graphic medium, and it identifies the building’ delimitation as a square grid. On the second
image, Gunzo, who is now free of his confinement to bed, meets Abbot Hugo to mediate the
project instructions. His unusual gesture is conspicuous: this is not a rhetorical gesture, as
Carolyn Carty proposes,20 but technical information that completes the message. The ges-
ture joins the breaths of the two hands to the length of the two stretched out thumbs, the
thumbs touching each other at their heads. Exactly in this way the early Renaissance archi-
tect Antonio Averlino, called “il Filarete”, illustrates and describes the foot measure around
A.D. 1460.21 This rule of thumb represents the foot length definition already specified by
Herodotus and mentioned by Gerbert dʼAurillac in the Middle Ages.22

20 Ibid.
21 Cf. Filarete, Libro di architettura (ca. 1465), Cod. Magliabechianus, Liber I fol. 4v.: “[...] Unaltra misura anco-
ra chiamata pie laquale pocho susa pure in alcuno luogho susa & questo pie e dimisura didue mani strette / o
vuoi dire raccolte lequattro dita el quinto disteso & agiugnierlo dipunta luno alaltro questo sidice essere un pie
[...]”; illustration “la misura del pie” on fol. 4r. (Cod. Magl., Liber I. Firenze, BCNF, Fondo Nazionale ms II.I.140.
[Spencer 1965, II (facsimile)].
22 Gerbert, De disciplinis mathematicis: Geometria II; Pl. 139, 96–98 (after Naredi-Rainer 1982, 105, n. 138).
 This fixed unit of length delimits the building and makes its dimensions rationally
comparable. It characterizes the rectangular cord grid of the central image as an arithmetic
scale system. Beyond this, it is easily imaginable, that Gunzo makes use of his gesture to ex-
press visibly the length of one perch or a multiple of it. He would demonstrate a rational scale
relation between a small preliminary visualisation and the real-size construction. In this way,
the conception of the aula imperialis, for example, could be presented in a few minutes. And
there is another text aspect to remark: “Nec de sumptibus dissidat”—the abbot is obliged not
to be repugnant to the construction costs. The document limits its information related to the
Cluny III project strictly to the high priority questions of any important decision in the build-
ing sector: demand analysis, dimensioning, costs, and project management. All this indicates
a real background of the story that uses the vision of saints as pia fraus to promote a giant
and expensive building project.

…inherited from Roman Antiquity

Looking for instruction placed at the disposal of master builders, masons, and geometers
around 1100 AD, there existed, besides a partial translation of Euclid, only the treatises of the
gromatici or agrimensores, called Corpus agrimensorum. The area of eastern France/Upper
Rhine was at that time a centre of circulation of this compendium.23 The Benedictine abbeys
like Cluny accomplished the document tradition and transcription. Indeed, all units of mea-
sure and related systems, as well as the square grid idea, observed in the mediaeval buildings,
are part of the content of the Corpus agrimensorum.24 The large scaled orthogonal coordinate
system, by which the modern eye identifies the agrimensores, was not applied in the High
Middle Ages. Instead, another important instruction: By word and image, the agrimensores
oppose an imaginary ideal geometry to the recommended realization, shaped after the na-
ture of the locality in question with existing structures (Fig. 15). The discrepancies between
a geometrical ideal and the reality observed in mediaeval buildings appear also on the urban
level, with the same characteristic graphic patterns. The recognition of the topography as an
important factor in mediaeval building practices, together with the repartition of dated edi-
fices, leads to completely new reconstructions. In Cluny, an approximately rectangular street
network completes two older, curved Gallo-Roman roads. The hitherto accepted polycentric
town evolution of Cluny had to be revised in favour of one single nucleus. Simple and logic
answers were found for concrete problems, e.g., why the main gate of the abbey is situated
so strangely aside at half height.25 The respect of the topography in mediaeval planning and
space organization, proposed by the Corpus agrimensorum, seems to be a general phenome-
non of the period: In Freiburg im Breisgau, the oldest known mediaeval town foundation in
Germany, the town perimeter and streets going back to the foundation period show the same
geometrical characteristics as the building plans analysed in Cluny. They were laid out after

23 Cf. Reeve 1986, 1–6.

24 Cf. Blume/Lachmann/Rudorff 1852, 355.
25 For elaboration and discussion of these and the following topics, see Flüge forthcoming, chapter VI (cf. n. 1
of the present article).
the topo-geographic conditions in three steps: first the town’s main axis, the cardo, parallel
to an older long distance path accompanying the foot of a Black Forest mountain from south
to north, then the town perimeter approaching a square contour around the main axis—
with deformations due to geographical relief and existing structures such as a few roads and
buildings, and thirdly the new part of the road system, strictly oriented after the drain direc-
tion of the terrain (Fig. 16).26 There exist already late Roman examples of the orientation of a
geodetic grid after the drain direction.27 The famous water runs of Freiburg have commonly
been dated into the years around 1170 after a series and concatenation of erroneous inter-
pretations of the accumulated archaeological street levels, beginning in Gauchstrasse, by the
local mediaeval archaeologists. The channel network can now be explicated in a new way
as an element and even, with the utmost probability, a condition of the original plan around
1100.28
Within high mediaeval town realizations, constitutional elements of the Roman space
organisation are preserved: as central institutions the crux viarum (the High Cross) and the
forum (market). The new observations establish that high mediaeval building and urban con-
ception and realization practices are in essential aspects closer connected to Roman Antiqui-
ty than the idealized post mediaeval ones.29

26 The Bertolds, later named Zähringer, Dukes of the Veronese Mark and Earls of Breisgau, were closely bound
to the Cluny Abbey. Her(i)mann(us) I., the brother of the foundator of Freiburg, Bertold II, was initially himself
Earl of Breisgau and Earl of the Veronese Mark, and he became later monk in Cluny, where he died in 1074 (cf.
Parlow 1999, XXVIII. The celebrated abbot Hugues de Semur (1049–1109 wrote a vita of Her(i)mann(us)).
27 Antique centuriation according to the draining direction existed e.g. near Padua, at the via Postumia (de-
cumanus) and the state road Padua-Bassano (cardo), as well as along the channel Carrara-Bovolenta-Pontelon-
go; cf. Galsterer 1992, 418.
28 Cf. Flüge forthcoming, Kap. VII.3.1.c.
29 L.c., Kap. VII.3.2.
Picture Credits
Hans-Peter Vieser: 1B, 9A–B.
Property of the author: 1A and C, 2A–C, 3A–B, 4, 5, 6A, 7, 8A–C, 10A–B, 11, 12, 16A–C.
Public domain due to copyright expiration: 6B (Representation of the cathedral domain in
the “Eadwine Psalter” (around 1155–1170, Cambridge, Trinity College, Ms R.17.1 fol.
285). Historical graphic rendering with unprecise legend reproduction [after Benev-
olo 1983, p. 186, Fig. 515, without signature]); 13 (B.n.F., Ms fr. 19093 fol. 14v); 14A
(B.n.F., Ms lat. 17716 fol. 43r); 14B (Firenze, BCNF, Fondo Nazionale ms II.I.140 (Co-
dex Magliabechianus) fol. 4r [“Su concessione del Ministero per i Beni e le Attività
Culturali / Biblioteca Nazionale Centrale di Firenze”, after Spencer 1965, facsimile]);
15 (B.A.V., Cod. Pal. Lat. 1564 c. 89v [after Guidoni 1981, p. 135, fig. 189]).

Bibliography
Benevolo, Leonardo 1983. Histoire de la ville (Roquevaire) [Storia della città. Rom/Bari 1975].
Carty, Carolyn M. 1988. “The Role of Gunzo’s Dream in the Building of Cluny III,” in Interna-
tional Center of Medieval Art (ed.), Gesta 27 (Michigan), 113–123.
Conant, Kenneth John 1968. Cluny. Les églises et la maison du chef dʼordre. Publication of the
Mediaeval Academy of America, Cambridge, MA (Mâcon).
Dinter, Peter (ed.) 1980. Liber tramitis aevi Odilonis abbatis, corpus consuetudinum monasti-
corum X (Siegburg).
Dauzat, Albert and Dubois, Jean and Mitterand, Henri 1971. Nouveau dictionnaire éty-
mologique et historique [Larousse] (Paris).
Flüge, Bernhard forthcoming. Domus solaratae der Periode Cluny III. Steinhaus und Stadtan-
lage um 1100 am Beispiel Cluny—Baugeschichtliche Grundlagen zur Erkundung des
Hochmittelalters (mit Beitrag zur europäischen Planungsgeschichte) (Max Planck In-
stitute for the History of Science, Studies 2; Berlin).
Flüge, Bernhard 2011. “Domus solaratae der Periode Cluny III. Licht-Lösungen an Profan-
bauten der Zeit um 1100,” in Schneider, Peter and Wulf-Rheidt, Ulrike (edd.), Licht—
Konzepte in der vormodernen Architektur (Diskussionen zur Archäologischen Baufor-
schung 10; Berlin), 289–303.
Galsterer, H. 1992. “Die Kolonisation der hohen Republik und die römische Feldmesskunst,”
in Behrends, Okko and Capogrossi-Colognesi, Luigi (edd.), Die römische Feldmess-
kunst. Interdisziplinäre Beiträge zu ihrer Bedeutung für die Zivilisationsgeschichte Roms
(Abhandlungen der Akademie der Wissenschaften zu Göttingen, Philologisch-His-
torische Klasse, Folge 3, 193; Göttingen), 412–428.
Guerreau, Alain 1998. “Espace social, espace symbolique : à Cluny au XIe siècle,” in Revel,
Jacques and Schmitt, Jean-Claude (edd.), Lʼogre historien. Autour de Jacques Le Goff
(Paris), 167–191.
Guidoni, Enrico 1981. La ville européenne. Formation et signification du quatrième au onz-
ième siècle (Brussels and Liège) [La città europea, 1978 (Milan)].
Hubert, Étienne 1990. Espace urbain et habitat à Rome du Xe à la fin du XIIIe siècle (École
française de Rome & Istituto storico italiano per il Medio Evo (edd.), Nuovi studi stori-
ci 7; Rome and Selci Umbro).
Méhu, Didier 2002. “Un monument de lʼhistoriogaphie clunisienne du XIIIe siècle : le manu-
scrit Paris, BnF, Latin 17716,” in Cluny ou la puissance des moines (Dossiers dʼarchéol-
ogie nº 269; Paris), 56.
Parlow, Ulrich 1999. Die Zähringer (Veröffentlichung der Kommission für geschichtliche
Landeskunde in Baden-Württemberg, Reihe A, Quellen, 50. Band; Stuttgart).
Reeve, Michael David 1986. “Agrimensores,” in Reynolds, Leighton Durham (ed.), Texts and
Transmission. A Survey of the Latin Classics (Festschrift to the 80th anniversary of Sir
Roger Mynors; 2nd edn.Oxford), 1–6.
von Naredi-Rainer, Paul 1982. Architektur und Harmonie. Zahl, Maß und Proportion in der
abendländischen Baukunst (Köln).
Wollasch, Joachim 1996. Cluny, Licht der Welt. Aufstieg und Niedergang der klösterlichen Ge-
meinschaft (Düsseldorf).

Archaeological records
Roiné, Nadine 1993. Cluny, Aménagement de la Place du Marché (ancienne cour des Écuries)
(Service Régional de l’Archéologie Dijon, Arch. record of the excavations from Dec.
1988, typescript in property of its author; Cluny).
Flüge, Bernhard forthcoming. Domus solaratae der Periode Cluny III, appendix, buildings
arch. documentation Pl. 1–50 (Max Planck Institute for the History of Science, Studies
2; Berlin).

Ancient sources
Antonio Averlino, detto il Filarete, Libro di architettura (ca. 1460), Firenze, BCNF, Fondo Na-
zionale ms II.I.140 (Codex Magliabechianus) fol. 4r.
Corpus agrimensorum, B.A.V., Cod. Pal. Lat. 1564, c. 89v.
Gilo, Vita Hugonis, Bibliothèque nationale de France, Ms lat. 17716 fol. 43r.
Villard de Honnecourt, Livre de portraiture (around 1220–1240), B.n.F., Ms fr. 19093 fol. 14v.

Edited sources
Blume, Friedrich and Lachmann, Karl and Rudorff, Adolf August Friedrich and Mommsen,
Theodor and Bursian, E. 1848/1852. Die Schriften der römischen Feldmesser ([Corpus
agrimensorum or] Gromatici veteres, vol. I/II). Vol. I, Karl Lachmann [texts] und Adolf
Rudorff [drawings] (edd.), Texte und Zeichnungen, 1848. Vol. II, Friedrich Blume, Karl
 Lachmann, Theodor Mommsen and Adolf Rudorff 1852, Erläuterungen und Indices
(Berlin).
Spencer, John Richard (ed. and transl.) 1965. Filareteʼs Treatise on Architecture. Codex Magli-
abechianus. (Yale Univ. Press; New Haven and London).
Gerbertus. De disciplinis mathematicis: Geometria II; Pl. 139, 96–98 [after Naredi-Rainer
1982].

Dendrochronological dating reports

Archéolabs réf. ARC 06/R3275D. Dormoy, Christian, Feb. 13th, 2006, Expertise dendrochro-
nologique dʼéchantillons provenant des Écuries de Saint-Hugues à Cluny (71250) (St-
Bonnet-de-Chavagne).
Archéolabs réf. ARC 06/3275D/2. Dormoy, Christian, Sept. 7th, 2006, Expertise dendrochro-
nologique dʼéchantillons provenant des Écuries de Saint-Hugues à Cluny (71250) (St-
Bonnet-de-Chavagne).
Archéolabs N/réf. ARC 05/R3325D. Dormoy, Christian, 2005, Expertise dendrochronologique
dʼéchantillons provenant du 23, rue Filaterie à Cluny (71250) (St-Bonnet-de-Chavagne).