This article was downloaded by: [University of West Florida]

On: 28 December 2014, At: 11:37

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41
Mortimer Street, London W1T 3JH, UK

The Art Bulletin

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/rcab20

Roman Labyrinth Mosaics and the Experience of Motion

Rebecca Molholt
Published online: 30 Jan 2014.

To cite this article: Rebecca Molholt (2011) Roman Labyrinth Mosaics and the Experience of Motion, The Art Bulletin, 93:3, 287-303, DOI:
10.1080/00043079.2011.10786009

To link to this article: http://dx.doi.org/10.1080/00043079.2011.10786009

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the
publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or
warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and
views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by
Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary
sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs,
expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with,
in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction,
redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly
forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
 Roman Labyrinth Mosaics and the Experience of
Motion
Rebecca Molholt
Only he who walks the road on foot learns the power that it
commands. . . .—Walter Benjamin, “Chinese Curios,” One-Way
Street, 1925–26
In a letter written in 1917, Walter Benjamin speculated on
the essential difference between painting and the other arts:
“From the human point of view, the level of drawing is
horizontal, that of painting, vertical. . . . A picture wants to be
held vertically before the viewer.”1 Benjamin had recently
been reading the art historian Alois Riegl, specifically his Late
Downloaded by [University of West Florida] at 11:37 28 December 2014
Roman Arts Industry, from which he adopted the notion that
works of art expressed their own sense of volition. As Benja-
min later developed the idea, the viewing needs of the paint-
ing implied vertical placement, in emulation of a window,
while other cultural forms—the architecture of the Paris
arcades, graphics and drawings, written texts—are better
seen as horizontal in nature, since they demand to be under-
stood as a “cut” through the “world substance,” a transverse
section in which some kind of line is envisioned to move
through and over the ground presupposed by the work. For
Benjamin, such horizontal experience entails a radically dif-
ferent set of perceptions, for, instead of being taken in all at
once, as a whole, the horizontal form is perceived in parts
whose import must be reconstructed imaginatively by the
viewer or reader. Fragments and discontinuity characterize
the horizontal, whereas integration and transparency distin-
guish the vertical. Benjamin recognized the way such vertical
orientation marked the modern epoch, from its art to the way
it conceived the study of history, but he wished to propose
other modes as alternatives to a modernism that had, by the
1920s, exhausted itself.
The study of Roman floor mosaics has traditionally been
encumbered by the kind of myopia Benjamin describes in his
distinction between vertical and horizontal forms of viewing:
modern scholars have tended to regard mosaics as if they
were paintings or were created in emulation of painting.2
The rise of easel painting in modern culture has installed an
unconscious privileging of this visual medium above all oth-
ers, with the result that the conditions of viewing attending a
painting hung on a wall have become normative for the arts
as a whole. The vertical bias has had several effects on our
understanding of floor mosaics. First, it has been assumed
that the aim of ancient mosaic designers was generally to
imitate painting, that mosaics are (or attempt to be) essen-
tially paintings in stone laid out on a floor. From this follows
the almost universal practice of installing mosaics on mu-
seum walls, in conformity with the presentation of paintings.
Conventions of Western perspective since the time of Leon
Battista Alberti understood that a painting was made by an
artist standing at a fixed point before it, and that a single,
immobile point near the center of the work is the implicit
location a viewer is supposed to reclaim in order to experi-
ence its spatial effects. This monocular perspective is repli-
cated in the kind of still photography that most often pre-
sents such works in textbooks and other published studies.
Scholars have complained that “[m]osaics are difficult to
photograph even under good conditions: because of their
size and situation, often only an oblique view is possible.”3
Measured by the standards of wall painting, the Roman
floor mosaic remains a fragment. Seen en face, from a fixed
position, or judged by the standards of realism established for
another medium, the mosaic is misread, and the proper
experience of its viewing unduly constrained. As Benjamin
might have put it, the floor mosaic wants to be regarded
horizontally, not vertically.
The meaning of the Roman floor mosaic was inseparable
from its experience as a tangible surface, one typically appre-
ciated by an ambulatory viewer situated in and aware of a
specific architectural setting. We need to rethink such mosa-
ics as forms and materials underfoot and to examine them
kinesthetically, as experiences that are by no means purely
visual. Footsteps can define a place— even an imaginary
place.4 The way that the labyrinth mosaics interact with the
architecture and activities of the baths of North Africa from
the late second to the early fourth century CE and with the
viewer’s ambulatory occupation of these spaces accounts for
the popularity of the theme in such settings. The traversal of
labyrinth mosaics in a bath context takes on a broader,
metaphoric meaning, since the pavements were deliberately
designed to blur the boundaries between life and myth.5
Myth offers a common realm available and accessible to all,
and the labyrinth’s iconography and mythography are partic-
ularly apposite in a bath context. As a floor decoration, the
labyrinth can reinterpret the space it defines, and walking
across these spaces helps to construct the bather as a heroic
athlete.
Roman North Africa was one of the wealthiest regions of
the empire in the second and third centuries CE, not least
thanks to its major exports of grain and olive oil; as a result,
these provinces are among the richest in surviving monu-
ments. The first emperor from North Africa, Septimius
Severus of Leptis Magna (in modern Libya), came to the
throne in 193, and his dynasty held power until 235. In the
provinces of North Africa, the great agonistic, Greek-style
athletic contests were extremely popular, especially during
the second, third, and fourth centuries,6 when the mosaics
and bath buildings discussed here were made. Games and
sporting events were staged on both the imperial and the
local levels, celebrating such municipal events as the dedica-
tion of bath buildings that still dot the landscape today.
Fifty-six known Roman floor mosaics represent the convo-
lutions of the labyrinth.7 While the majority come from
houses, fourteen labyrinth mosaics come from baths, and
seven of these from baths in North Africa.8 Stories involving
 288 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
Downloaded by [University of West Florida] at 11:37 28 December 2014
the labyrinth appear in sources ranging from Herodotus to
Pliny the Elder, Plutarch, and Apollodorus.9 Many narratives
intersect at this motif: the pride of Minos, the lust of Pasi-
phae, the love of Ariadne, the cleverness of Daedalus, the
tragedy of Icarus, the horror of the Minotaur, and the hero-
ism of Theseus. But the bath mosaics highlight the athletic
prowess of the hero Theseus.
The popularity of labyrinth mosaics has been ascribed
variously to their gamelike aspect,10 their apotropaic na-
ture,11 their near-abstract design that allows for infinite ex-
tension across a room of any size,12 their ability to reflect an
elite Roman cultural status on viewers having the education
to recognize the image and its multiple ramifications,13 their
ability to represent to the Roman viewer the triumph of
civilization over barbarism, and their image as a representa-
tion of the city itself.14 Yet, though it has been noted several
times,15 no one has explained why the labyrinth mosaics were
such a common theme in the baths of North Africa. Of the
sixteen labyrinth mosaics in North Africa, seven adorned
baths, but these images and their contexts have never been
studied together.16
Walking on the Labyrinth: Surface and Traversal
Floor mosaics require the beholder to think on his or her
feet. Any pavement will probably be touched (even felt by
unshod feet) at the same time that it is seen. This matters
because the Greeks and Romans believed that vision itself was
both haptic and optic. By a process of extromission, the eyes
released rays that traveled out to touch the object in question
and then came back to the eyes.17 Visuality was thus not
anchored in the retina alone. Floor mosaics that imitate
painting are not necessarily playing to the strengths of the
mosaic medium; any illusion of deep perspective will always
be mitigated by the pavement’s tangible flatness and texture
underfoot. Images presented on a surface with which the
viewer has physical contact prompt the spectator at all times
to acknowledge personal involvement with the creation of
meaning. Especially effective in this light are mosaic com-
positions that endeavor to take into account the simulta-
1 Labyrinth mosaic, seen from the
eastern corner of the room, early 4th
century CE, frigidarium, Baths of
Theseus and the Minotaur, Belalis
Maior, Tunisia (artwork in the public
domain; photograph from Mahjoubi,
Recherches, fig. 87)
neity— of horizontality and of verticality— built into any
experience of a floor mosaic, where the image and any accom-
panying narrative are deployed at right angles to the standing
viewer.
The schema and scale of the labyrinth story presented in
Roman baths are specific to the medium of floor mosaics.
Wall paintings of Theseus and the Minotaur, such as those
surviving from the Vesuvian region, focus on the hero’s vic-
torious exit from the labyrinth: we see him outside.18 But the
mosaics depict a wholly different moment in the narrative
and spread out the patterns and convolutions of the labyrinth
itself (Fig. 1). Such pavements, stretching across large rooms,
rarely feature any of the ancillary characters (such as Ariadne
and the crowds of onlookers and grateful Athenian children)
who are often present in wall paintings. Instead, the floor
mosaics showcase the architecture of the labyrinth and pre-
sent a journey underfoot.
Because their narrative of journey unfolds across the sur-
face of a floor, labyrinth mosaics are representations of spa-
tial experience that unify art and architecture. The otherwise
unremarkable action of walking cannot be taken for granted
across these surfaces, since the mosaics call for immersion
and immediacy on the part of the observer.19 The experience
of mosaics thus involves a sort of phenomenological vision,
prompting a larger cognitive, perceptual, retinal, and episte-
mological effort toward understanding. Maurice Merleau-
Ponty’s emphasis on the “lived perspective of the visible world
in relation to our living body” provides an important model
here, since he describes perception as “our kinaesthetic,
prescientific lived-bodily presence to the world.”20 As she
actually treads on the images, the beholder is moved to
become actively engaged in the narrative unfolding under-
foot.
Roman baths are environments designed to serve and cel-
ebrate the human body; these are spaces created for rituals of
great physicality. Observers experienced the labyrinth mosa-
ics while also partaking of the social rituals of bathing. Bath-
ers moved through different temperature zones, in and out
of water of various temperatures, walked barefoot across cold
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION
floors as well as room-temperature floors and other chambers
with floors too hot to traverse without shoes. A heightened
bodily awareness was stimulated by the highlighted physical-
ity of the bathing experience itself. Some rooms would have
been full of steam and dim light, others full of sunlight and
the splash of water. A poem from the Latin Anthology even
links the bodily pleasures of bathing and the mental pleasure
provided by the decor of the bath itself.21
Simply as a dynamic pattern of tight rhythmic lines, laby-
rinths are a pleasing visual decoration. The contrast between
a violent and dangerous struggle (Theseus versus the Mino-
taur) and the pleasures of bathing was probably not lost on a
Roman audience.22 Because the sea god Neptune clearly
oversaw the watery realm of the bathhouse, a viewer con-
fronted with a labyrinth mosaic might well remember the
connection of Neptune to the Minotaur.23 Katherine Dun-
babin has suggested that “the remarkable popularity of laby-
rinth designs for the decoration of rooms in or attached to
Downloaded by [University of West Florida] at 11:37 28 December 2014
baths, though perhaps in part intended for the entertain-
ment of bathers who can puzzle out the maze while they
relax, may also have been influenced by the notion that the
wanderings of the maze baffle or distract the evil spirits or the
malevolent gaze.”24 Certainly, apotropaic resonance can be
accepted as a reason for the popularity of this myth in this
context.25
More nuanced interpretations can be made, however, from
analyses of labyrinths, taking into account their different
settings in the bath, their relation to the surrounding archi-
tecture, and the imagery of the rest of the bathing ensemble,
such as sculptural programs. Furthermore, the labyrinth mo-
saics make fresh sense when we situate them within their
architectural settings. These pavements, themselves illustrat-
ing a journey, rely also on the physical movement of the
beholder; while the realm of the mosaic begins at the en-
trance to the room, only an oblique view of the entire com-
position is available from that vantage point. The narrative
will not culminate until one steps into and then through the
room.26
Labyrinth mosaics often present both a picture and a pic-
ture of a plan, as the mosaic from the Baths of Belalis Maior
demonstrates (Figs. 1–3).27 The mosaic is still in situ, paving
an unheated room in a small thermal establishment from the
early fourth century. The labyrinth itself measures approxi-
mately 53⁄4 square yards (4.8 square meters).28 The outer part
of the mosaic represents heavy masonry walls splaying out on
all four sides, as if to hold the rest of the structure within the
high circuit of the depicted walls reaching to the very walls of
the room itself. A crenellated gate, one on each side of the
room, punctuates each mosaic wall, represented as huge
stone blocks laid in double courses. Only one gate, at bottom
right, opens to reveal a single arched doorway in the heavy
exterior wall offering access to the labyrinth’s interior. This
gate, though facing inward toward the labyrinth, corre-
sponded with the actual opening into the room. The laby-
rinth would first be seen from the doorway of the room, by a
beholder poised, perhaps between columns, looking into the
room’s southern half. The central image, however distant
and oblique, would be oriented toward the oncoming viewer.
Within the fortress walls are walls again, but now they make
no pretense to three-dimensionality and are rendered only as
289
2 Combat between Theseus and the Minotaur, early 4th
century CE, detail of center of labyrinth mosaic, frigidarium,
Baths of Theseus and the Minotaur, Belalis Maior, Tunisia
(artwork in the public domain; photograph from Mahjoubi,
“Le thème du Labyrinthe,” fig. 3)
lines, black and red on white. It is a floor plan of the convo-
lutions of a labyrinth, rendered on the floor of the Roman
frigidarium, a cold room that is the traditional finishing point
in the sequence of Roman bathing. Black and red lines
representing walls on the Belalis mosaic create and traverse a
winding pattern. Theseus finds the Minotaur in the laby-
rinth’s innermost chamber, shown at the center of the mosaic
and also the center of the frigidarium (note the ball of thread
between Theseus’s knees, Fig. 2). Even though beholders
probably did not bother to follow the tortuous lines of the
hero’s path, they would have been aware of Theseus’s mon-
umental journey underfoot.
It is rare that entire large-scale mosaic floors bespeak a
single narrative, though the labyrinth floors often do. More
often, patterns and sets of figural scenes are collaged to-
gether. Multiple borders often frame mosaic pictures (em-
blemata), and these borders regularly tell their own ancillary
stories.29 Even the massive floors of North African baths that
are transformed by mosaics into spectacular and continuous
expanses of the ocean show episodes from many different
aquatic narratives taking place across the realm of the
room.30 But the image within a labyrinth is not isolated by its
immediate mosaic enframement; the floor renders a single
myth, and the contextualized emblema at its center shows the
combat between hero and monster in the labyrinth’s inner-
most chamber.31
While the surrounding labyrinth is designed as if observed
from overhead and in plan view, the combatants in the center
are seen from the side and modeled to evoke three dimen-
sions (Fig. 2). At the center of the floor mosaic, where
Theseus and the Minotaur fill the available space, the tes-
serae are smaller and more colors are used. Because the
fortress walls reach to the very walls of the room itself, the flat,
horizontal floor mosaic reiterates the space of the room itself.
In a dialogue of art and architecture, depicted walls meet
actual walls; the very center of the labyrinth is also the center
of the frigidarium. The viewer’s movement across the space
of the labyrinth provides the narrative links and thus the
sense of common ground.
 290 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
Downloaded by [University of West Florida] at 11:37 28 December 2014
The play between two and three dimensions emerges as
one aspect of this game, but it is not the one mentioned by
the ancient sources. Pliny the Elder focused on a different
kind of spatial complexity: the compression of extended
space into a small area. The labyrinth, he wrote, contains
passages “that wind, advance and retreat in a bewilderingly
intricate manner.”32 While Pliny’s description highlights the
role played by multiple doors to hinder physical navigation, it
is a different confusion that accompanies the experience of a
labyrinth mosaic. The floor itself is a wide, flat expanse, easy
for any walker to traverse. But space is so visually compressed
within the mosaic lines of the labyrinth floor that it is inevi-
tably difficult (or even impossible) to follow any one path-
way—with the eyes. And so, the spectator is optically lost,
even though he is standing on an open floor. At Belalis
Maior, the visual instability is heightened by the alternations
of the red and black lines that compose the labyrinth, which
shimmer and shift back and forth at every turn, blurring the
distinction between figure and ground.
Scholars have wondered about the disjunction between the
floor mosaics displaying unicursal labyrinths—those that of-
fer only a single route forward, where in fact it would be
completely impossible to lose your way—and the literary
implications of a labyrinth as a maze in which one actually
could get lost.33 These theories imagine the labyrinth as if
scaled down and on paper; with a pen in hand, one surely
could trace a unicursal path from beginning to end. By
3 Plan of labyrinth mosaic, early 4th
century CE, frigidarium, Baths of
Theseus and the Minotaur, Belalis
Maior, Tunisia (plan by Mahjoubi, “Le
thème du Labyrinthe,” fig. 2)
contrast, Roman labyrinth mosaics are not mazes to be fol-
lowed physically. Indeed, only one (at Mactar, see below) is
large enough for a person to follow its course, albeit in tight
footsteps. These are instead visual mazes, and visually, across
the floor of a Roman bathhouse, it is virtually impossible for
the eye to stay the dizzying course. In every case, however, the
beholder has only to take a few easy strides over the flat
mosaic surface to reach the center of the labyrinth, which is
usually also the center of the room. A terrible journey
through high-walled corridors with no end in sight is com-
pressed to a pattern of flat lines, and the center is gained
easily. The story of Theseus and the Minotaur is, of course, a
well-trodden path with positive outcomes. The frame of these
mosaics, designed to imitate an enclosure of towering city
walls, helps to simulate the deep view underfoot, but the epic
journey is accomplished in abbreviated form, and the mythic
protagonist easily handles the danger at the center. Nonethe-
less, the labyrinth design invites all who enter this space to
give first-person attention to their movement, not just to walk
unthinkingly across the monumental passages condensed un-
derfoot.
Baths of the Labyrinth, Thuburbo Maius: Theseus as
Wrestler
For too long, the study of labyrinth mosaics has focused
exclusively on comparisons made within the mosaic corpus,
and in nearly every case the study has remained divorced
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION 291

from any consideration of a broader architectural and social

context.34 The size and dimensions of the labyrinth’s picto-
rial organization, so uncanny in book illustrations, nonethe-
less make perfect sense for the actual turning and moving
viewers on the mosaic, enmeshed in the story told on its
multiply oriented surface. The myth of Theseus and the
Minotaur gains new resonance when it is placed on the floor,
in the path and as the path of its beholder in the Roman
baths, a space designed as much for athletic activities as for
cleansing and relaxing. Theseus, after all, did defeat the
Minotaur in a wrestling match, and the hero is credited with
laying down the rules of this sport.35 Though Hercules and
Hermes are more commonly considered the patron gods of
sporting events, the heroic combat of Theseus and the Mi-
notaur was itself an athletic spectacle, and the pleasures of
Roman baths certainly included sporting activities. In his
otherwise magisterial 1977 study of labyrinth mosaics, Wiktor
4 Mosaic of boxers, late 3rd– early 4th century CE, from the
Daszewski erred when he ignored the “caractère sportif” of tepidarium, Baths of the Labyrinth, Thuburbo Maius, Tunisia.
Downloaded by [University of West Florida] at 11:37 28 December 2014

the Roman baths.36 The Bardo Museum, Tunis (artwork in the public domain;
The majority of large-scale African baths contained facili- photograph by the author)
ties for exercise and combat sport: palaestrae surrounded by
colonnaded courtyards.37 But smaller bathing establishments
borrowed the agonistic imagery, even if they might not have
offered exercise space within their walls. The high concen-
tration of agonistic iconography in Roman baths signifies the
sporting ambiance that dominated these buildings.38
Many bathers exercised before bathing.39 We know from
Roman authors—including Martial, Juvenal, and Seneca—
that exercises such as weight lifting, ball play, and wrestling
took place in the baths as a prelude to cleansing.40 Greek-
style combat competitions (boxing, wrestling, athletic games)
gained in popularity in proconsular Africa after the com-
mencement of the Pythian Games in Carthage in the Severan
era, as is attested by a rich epigraphic and iconographic
record.41 Later sources, too, are perfectly clear on this point:
a statue base from 378 CE from Sabratha (Libya), for exam-
ple, lauds Flavius Vivius Benedictus for restoring the local
baths and thereby restoring exercise to the people.42
At the small Baths of the Labyrinth at Thuburbo Maius,
late third to early fourth century, a mosaic of boxers deco-
rated the center of the tepidarium floor (Fig. 4).43 Both
boxers have their hands protected by tightly wrapped cestes, a
Roman form of boxing gloves. While the younger man re-
mains standing at right, the older man crouches in a defen-
sive posture, bleeding after a heavy blow to the head. Inscrip-
tions at other thermal establishments in the region tell us that
boxing combats were offered, sometimes when the bath was
dedicated.44 The Baths of the Labyrinth, barely more than
480 square yards (400 square meters) in size, lacked a palaes-
tra, though archaeologists speculate that a basilical hall at the
northwest may have fulfilled a sporting function.45 5 Labyrinth mosaic, late 3rd– early 4th century CE,
Whether or not people actually exercised within these frigidarium, Baths of the Labyrinth, Thuburbo Maius, Tunisia.
walls, the overall decor of the Baths of the Labyrinth aimed The Bardo Museum, Tunis (artwork in the public domain;
for an athletic ambiance, and I believe that the labyrinth photograph by the author)
mosaic adorning the adjacent frigidarium was another impor-
tant part of this milieu (Fig. 5). The frigidarium itself is
approximately 36 square yards (30 square meters) with two
steps leading down to a single cold-water pool in one corner. scintillating, shimmering surface when seen underwater. In
The pool was paved with large tesserae in different colors laid the normal course of events, the labyrinth and the frigi-
in a random pattern, and this mosaic would have presented a darium would be entered both before and after the beholder
 292 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
6 Drawings of the same geometric mosaic from the market at
Roman Hippo Regius, Algeria, contrasting perspectives of two
viewpoints: an oblique view and one directly en face (drawings
by Lassus, reproduced with permission from Lassus, “La
mosaı̈que romaine,” pl. CLIX, fig. 1)
Downloaded by [University of West Florida] at 11:37 28 December 2014
had spent some time contemplating and traversing the mo-
saic of boxers in the tepidarium.46
Although the labyrinth mosaic paving the floor of the
frigidarium has been damaged, an interesting array of archi-
tectural structures still serves as its border: heavy masonry
facades and black doorways seem to open onto cold, dark
recesses beyond. Blank walls and multiple entrances must
originally have gone completely around the edges of the
Thuburbo labyrinth mosaic, presenting a host of gates and
openings to the viewer enclosed within.
All these open doors recall Pliny the Elder’s description of
an Egyptian labyrinth:
[It was] quite the most abnormal achievement on which
man has spent his resources. [. . . This Egyptian labyrinth
was the model for the one Daedalus built on Crete] con-
taining passages that wind, advance and retreat in a bewil-
deringly intricate manner. It is not just a narrow strip of
ground comprising many miles of “walks” or “rides” such
as we see exemplified in our tessellated floors or in the
ceremonial game played by our boys in the Campus Mar-
tius, but doors are let into the walls at frequent intervals to
suggest deceptively the way ahead, and to force the visitor
to go back upon the very same tracks that he has already
followed in his wanderings. [. . . All are] alike in being
roofed with vaults of carefully worked stone. There is a
feature of the Egyptian labyrinth which I for my part find
surprising, namely an entrance and columns made of
Parian marble. The rest of the structure is of Aswan gran-
ite, the great blocks of which have been laid in such a way
that even the lapse of the centuries cannot destroy them.47
With the multiple angles of viewing set up through the
depicted architecture, the surface of the mosaic ceases to
be a fixed and static picture and turns instead into a
structure within which the viewer must circle around to see
it from all sides and must look out at the multiple axes of
the walls from within. Even a simple pattern (like that of
the labyrinth’s black lines across the white ground) ap-
pears notably more dramatic and dynamic when seen from
an oblique angle (Fig. 6).48
As at Belalis Maior, the heavy architectural exterior of the
Thuburbo labyrinth mosaic gives way to an orderly delinea-
tion of the labyrinth’s interior walls. Again, the endlessly
twisting hallway winds through all four quadrants of the
labyrinth before opening onto the square field at the center
(Fig. 7), where Theseus and the Minotaur appear amid the
strewn human remains of the monster’s former victims: a
head, a severed arm, a single foot, and what appears to be a
leg bone bracket the combatants. On a sand-colored surface,
Theseus lunges inward to kick or knee the Minotaur’s flank.
The Minotaur, as Apollodorus said, had “the face of a bull,
but the rest of him was human.”49 On one knee, the monster
appears entirely at the hero’s mercy. Theseus has wrenched
back his head by one horn, and though the Minotaur grabs
Theseus’s elbow in futile protest, it is obvious that a blow
from the hero’s curved stick (pedum) is about to dispatch the
monster.
So far, there has been nothing to suggest this mosaic
presents anything but a mythic, epic combat. But several
details actually blur the line between sporting events held in
the mythic and worldly realms. In the Roman context, the
heavy athletic disciplines were wrestling, boxing, and the
pankration, a brutal blend of wrestling and boxing. There
were no weight classes in antiquity, and so these sports were
not for small men. The advantage went to the heavier man in
the struggle to throw his opponent off balance and to the
ground.50 The boxers in the next room at Thuburbo Maius
are both heavy, bulky men, and within the labyrinth, Theseus
and the Minotaur are no lightweights, either. Both man and
monster have similar physiques—thick arms and legs and
matching pot bellies. Dio Chrysostom disparaged wrestlers as
“pot-bellied bullies,” and in the second century Galen criti-
cized them for their long meals and their practice of force-
feeding themselves, all in order to get their weight up.51 On
the Thuburbo mosaic, the decidedly human, nonheroic body
type of Theseus blurs the boundary between mythic and
everyday combatants.52
The fighting strategy employed by Theseus in the central
scene is also copied from observations of contemporary sport-
ing events like the pankration. “Turn your body sideways to
your opponent and grip him by the head with your right
hand,” directs a notation from a papyrus found in the Egyp-
tian city of Oxyrhynchus, giving instructions for wrestling
practice.53 Though Theseus reserves his right hand for his
pedum, he clearly has a firm grip on one horn of the immo-
bilized Minotaur.
Plutarch describes the joy of King Minos when Theseus
overcame the Cretan bull, father of the Minotaur, in a wres-
tling match at funeral games. Like Ariadne, who was watch-
ing, “Minos also was delighted with him, especially because
he conquered Taurus in wrestling and disgraced him.”54
Apollodorus tells us that Theseus killed the Minotaur by
smiting him with his fists, and Plutarch reveals that the hero’s
journey into the labyrinth was undertaken “carrying no war-
like weapon.”55 The pedum was Theseus’s only weapon be-
sides his skill at fighting and his not inconsiderable bulk and
strength.56
The Minotaur’s pose would have caused any Roman ref-
eree to yell out “Round over!” for touching one knee to the
ground signaled a loss in the wrestling arena.57 Rather than
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION 293
Downloaded by [University of West Florida] at 11:37 28 December 2014

7 Combat between Theseus and the

Minotaur, late 3rd– early 4th century
CE, detail of center of labyrinth
mosaic, from Baths of the Labyrinth,
Thuburbo Maius, Tunisia. The Bardo
Museum, Tunis (artwork in the public
domain; photograph by the author)

8 Four young athletes with classic

short wrestling haircuts, each wearing
a necklace, three with amuletic
pendants; the two pairs flank a table
bearing palms of victory and a prize
crown, late 3rd– early 4th century CE,
from a threshold in the House of the
Boxers at Utica, Tunisia, mosaic. The
Bardo Museum, Tunis (artwork in the
public domain; photograph by the
author)

showing us an evenly matched pair of combatants at the
outset of the contest, mosaic battles of Theseus and the
Minotaur regularly show the man vanquishing the beast in a
victory clearly marked by the pose of the Minotaur with his
knee to the ground. This can be seen at Belalis Maior as well
and on labyrinth mosaics from Roman sites around the Med-
iterranean.58
Other aspects of the Thuburbo mosaic find further paral-
lels with everyday Roman sporting practice, including The-
seus’s spiky, closely cropped hairstyle. Wrestlers, especially
professionals, generally wore their hair short so as not to offer
their opponents any long hair to grip and pull.59 Further-
more, Theseus is shown wearing a bulla or amulet on a
red-brown cord around his neck, as are all four of the young
boxers or wrestlers on a mosaic from Utica (modern Tuni-
sia), late third to early fourth century (Fig. 8), as does the
bath attendant labeled “Tite” in a mosaic from the late Ro-
man villa at Piazza Armerina, Sicily.60 Regular bath goers,
who doubtless came in all shapes and sizes, might well be
undressed like Theseus and wearing such an amulet accord-
ing to contemporary practice.61
As a matter of course in images of athletic contests, the
older opponent tends to be vanquished by the younger. This
happens in the boxing scene in the tepidarium at Thuburbo
 294 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
Downloaded by [University of West Florida] at 11:37 28 December 2014
9 Labyrinth mosaic seen from the west, 150 –200 CE,
frigidarium, baths at Hippo Regius, Algeria (artwork in the
public domain; photograph by Marcel Bovis, reproduced with
permission from Lassus, “La mosaı̈que romaine,” pl. CLIX, fig.
2)
Maius, where the bearded, older man is shown bleeding, the
younger man upright and balanced on his toes after scoring
a major hit (Fig. 4). While the boxers in the next room at
Thuburbo Maius presented a blunt and even brutal portrait
of human athletics, the wrestling match of Theseus and the
Minotaur was of more heroic and mythical proportions. The-
seus the athlete goes far toward explaining the popularity of
the labyrinth in bath contexts. Hercules was also shown em-
ploying the poses and holds commonly used in the palaestra
in epic wrestling matches against Antaeus, Achelous, Triton,
and even the Nemean lion.62 We should add Theseus to this
category of mythical-cum-Roman sportsmen.
Baths at Hippo Regius: Minotaur as Athlete
Mosaics of the labyrinth, depicted at a large scale in the realm
of the Roman bath, and by extension their viewers acquire
the semiotic status of performativity. Any image at the center
of the labyrinth (and the center of the room) would not be
legible at first glance, so a viewer looking in from any door-
way would feel impelled to explore and traverse the labyrinth,
like Theseus. The physical action of crossing these floors
quickly embeds the observer within a mythic narrative.
The frigidarium of a small private bath in the center of
Hippo Regius (in modern-day Algeria) must have been a
spectacular place. Its construction and mosaics date from
150 –200,63 and expensive slabs of marble covered the frigi-
darium walls. The cold plunge was also lined with marble and
flanked by marble niches in which bathers could take their
leisure. A massive black-and-white labyrinth mosaic paves the
10 Drawing of labyrinth mosaic, 150 –200 CE, frigidarium,
baths at Hippo Regius, Algeria (plan by E. Stawski, reproduced
with permission from Marec, “Le thème du labyrinthe,” pl.
CCIX, fig. 2)
frigidarium floor, a space measuring nearly 7 feet 8 inches by
6 feet 7 inches (7 by 6 meters) (Fig. 9).64 Again, the mosaic
displays a scene that must be experienced to be understood;
the image on the far side would become visible only after
several steps had been taken. The mosaic image would trans-
port the viewer optically, even as his body was also in motion.
Heavy black walls, indicated in the mosaic as four courses
high and topped by a rippling line of crenellations, entirely
surround this labyrinth (Fig. 10). On the north side of the
room, the fictive walls are pierced by a single gate through
which a thick black line wends its way through the labyrinth’s
interior. This line is the path taken by the thread of Ariadne,
who fell in love with Theseus when she saw him in the
wrestling match.65 Ariadne gave Theseus a ball of thread, one
end of which he was to fasten to the lintel of the labyrinth
door. Holding the ball in his hand, he was to unwind it while
penetrating deeper and deeper into the labyrinth. The
thread was not shown navigating the Thuburbo Maius mosaic
(Fig. 7), and, while the labyrinth at Belalis Maior was de-
signed in alternating lines of red and black, the ball of thread
was featured only at the conclusion of the narrative in the
central scene (Fig. 2). But here, at Hippo Regius, the thick
line of the thread weaves throughout the entire room, be-
tween the thinner lines representing the labyrinth walls.66
The single, wide line of the thread and the multiple, thin
lines of the walls appear serrated because the tesserae are set
diagonally across the floor. This jagged line breaks up the
monotony of the pattern and gives tautness to the separate
elements.
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION
Downloaded by [University of West Florida] at 11:37 28 December 2014
11 Minotaur, 150 –200 CE, detail of the center of the labyrinth
mosaic, frigidarium, baths at Hippo Regius, Algeria (artwork
in the public domain; photograph from Marec, Hippone, fig.
54)
The path of the thread culminates in the center with the
large black ball of thread punching into the middle zone
(Fig. 11). Here, a torso-length portrait of the Minotaur, ren-
dered in gray, black, and white, already occupies this square
area. Long eyelashes, pointed ears, a thickly muscled chest,
and curving horns are all shown in some detail, and his head
is turned slightly, almost coyly, away from the intruding ball
of thread. The Minotaur’s image recalls nothing so much as
the bust-length mosaic portraits of heavyset athletes from the
frigidarium of the Baths at Thapsus, the Antonine Baths in
Carthage, and the exedrae of the Baths of Caracalla in
Rome.67
At the small baths of Hippo Regius, this labyrinth mosaic
has already been successfully traversed: the thread, borne
through many winding corridors, has arrived at the center.
Theseus, however, is nowhere to be seen, and the role of the
hero is left to the beholder, who has also crossed the laby-
rinth. This gives us a clear instance of what Wolfgang Iser
would have called a “constitutive blank.”68 What is missing
from the scene (the hero/protagonist) prompts the viewer to
accord special attention to his own status. The viewer has
already assumed all the intermediary movement and expen-
diture of time and easily accomplished the laborious journey
to the center of the labyrinth. When both Theseus and the
Minotaur are represented, as at Belalis Maior and Thuburbo
Maius, the mosaic displays not an equal contest but a victory.
No labyrinth mosaics ever show Theseus alone at the center,
but at Hippo Regius, the Minotaur appears like an opponent,
ready for the match and all comers.69
As Alois Riegl has written, “every work of art does presup-
pose the existence of a perceiving subject,”70 and here the
beholder has stepped into the role of the hero since his own
295
12 Drawing of a now-lost labyrinth mosaic with a helmet at its
center, before 79 CE, from Pompeii, Italy (artwork in the
public domain; photograph from Gli ornati delle pareti e i
pavimenti delle stanze dell’antica Pompeii incisi in rame [Naples:
Stamperia Regale, 1808], pl. 93)
movement, both visual and physical, implicates him in the
myth underfoot. The idea of mobility is already present in the
work, as the arrival of the thread at the center demonstrates
that a journey through the labyrinth has been accomplished.
The Minotaur, at the center, must be mastered just as the
labyrinth has been mastered. The Hippo Regius labyrinth
transforms the motif from a visual game to something more
like a physical event. Because the observer is situated within
the setting of the labyrinth, he gains the status of a character
in the drama underfoot, traveling within and along the
course that must be experienced and apprehended.
The path for the feet alters what had been merely a game
for the eyes. As the viewer moves from edge to center and the
myriad paths underfoot are stepped on and over, great dis-
tances for the eye and the imagination are traversed in a few
steps. It is the beholder who arrives at the center of the
labyrinth, to be confronted by the representation of an ad-
versary of mythic proportions. The heroic protagonist is not
imaged, but the hero can still be understood as present, now
in the body of the beholder.
A now-lost labyrinth mosaic from Pompeii offers an inter-
esting early parallel to this process of interactive viewing (Fig.
12). Rather than a portrait of the Minotaur, that labyrinth
had a helmet at its center, a point of confusion for every
commentator.71 The situation becomes clearer when we
think of Lucian’s description of the method of drawing lots
for the wrestling matchups at Olympia, which E. Norman
Gardiner explained in 1905: “Lots marked in pairs with the
letters of the alphabet in succession and corresponding to the
number of competitors were thrown into a silver helmet
sacred to that purpose from which each competitor in turn
 296 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
Downloaded by [University of West Florida] at 11:37 28 December 2014
drew a letter.”72 In the same way, the Pompeian mosaic
intimates an opponent waiting at the center of the mosaic:
the as-yet-unknown adversary’s name will be drawn from the
helmet.73 Again, the path across the mosaic stage embeds the
beholder within the larger narrative and, again, in confron-
tation with an adversary.
The Great East Baths, Mactar: Path to the Palaestra
Just as mosaics are not freestanding works of art independent
of the surrounding architecture, so their perception must
also include the architecture the mosaics simultaneously de-
fine as a physical surface and reinvent with their visual imag-
ery. Across the terrain of these labyrinth floors, it is the
perceiving subject who finds, in the flat floor underfoot, a
stage for a narrative that elevates his own movement and an
invitation to synchronize his actions with the work of art.
Visual perception is just one layer of this process of bodily
perception and projection: depicted space complements and
expands actual space via the introduction of history, imagi-
nation, and mythology. Motion is, once again, the key to the
labyrinth mosaic of the Great East Baths at Mactar, and here
again we encounter the unity of the work of art with the
architecture and an invitation to interactive viewing.74
Monumental, symmetrically planned, and dated to 199
by a dedicatory inscription, the establishment at Mactar is
classed as an “imperial type” of thermae (Fig. 13).75 Per-
haps the best-preserved Roman bathing edifice in North
Africa, it retains many walls that rise to a height of several
meters and often give an indication of the cross-vaulting
that once soared overhead. Two internal palaestrae are
surrounded on three sides by huge, U-shaped cross-vaulted
ambulatories. The palaestrae are set on either side of the
swimming pool, which opened to the countryside through
five large windows.76
Adjacent to each palaestra is a semicircular exedra. The
13 Plan of the Great East Baths,
Mactar, Tunisia, baths dedicated 199
CE, showing location of labyrinth
mosaic in northern hemicycle next to
palaestra (plan © Yegül, Baths and
Bathing, fig. 212, adapted by James M.
Morris)
floor of the northern hemicycle is entirely filled with a laby-
rinth mosaic in black, gray, and white (Fig. 14). The space of
the exedra is a kind of pivot for circulation in various direc-
tions, which recalls the multiple curves of the labyrinth itself
as well as Plutarch’s description of its “intricacies.”77 The
exedra can be entered from five doors within its curved walls,
and it gives onto the palaestra itself between six arcaded
piers.
Only the thread of Ariadne delineates the wide paths of
this mosaic labyrinth, which entirely fills the exedra. The
thread begins to unwind from the central doorway at the
back of the arc. The labyrinth culminates in a half circle
abutting the marble lintel of the central door on the
opposite side, which opens into the palaestra. Here the
rigid line of thread breaks into loose curves, as if to show
the last few feet unwound before the hero dropped the ball
of thread on the floor to commence his battle with the
Minotaur (Fig. 15).
At Mactar, the Minotaur is not shown, and neither is The-
seus. We are left at the threshold with nothing but an image
of a ball of thread on the ground and the insinuation of a
larger, missing totality. Rather than rendering recessional
space, this floor mosaic borrows from the space of the room
to generate its illusions of three-dimensionality, projecting a
winding image that claims to obey the same laws of gravity
that govern the viewer. Such an “addition” to real space
creates an ambient for arriving bathers and wrestlers to oc-
cupy and a space where the stage is set for narrative, but the
main characters are unseen, or at least unseen in the art. The
viewer’s entrance into the performance space outlined by the
mosaic is, on one level, manifestly possible. The high socles
or believable architectural forms that often occupy the fore-
grounds of Roman wall paintings are missing, as are the high
walls that surround other mosaic labyrinths. The floor itself is
bounded by high walls and multiple doorways, and the mo-
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION 297
Downloaded by [University of West Florida] at 11:37 28 December 2014

14 Labyrinth mosaic and palaestra, in

the Great East Baths, Mactar, Tunisia,
dedicated 199 CE (artwork in the
public domain; photograph by the
author)

saic has usurped this entire space of considerable dimensions

to evoke one story line of athletic heroism. At Mactar, the
here and now can never be detached from this mosaic nar-
rative.
One more step from the center of the Mactar labyrinth—in
fact, the inevitable next step forward—puts the viewer on the
threshold of the palaestra. Then a step down, and the be-
holder would gain the sand of the wrestling ring itself. Fol-
lowing a passage from Lucian, Ranon Katzoff posits that the
skamma (wrestling ring) should be understood as “a small
marked off terrace, perhaps somewhat raised or lowered, in
the court, where the sand was piled for sand wrestling.”78
Rather than providing a self-enclosed mythological narra-
tive already populated by a full cast of characters, the Mactar
mosaic is not an autonomous work of art. Following Riegl, the
external unity between image and viewer can be understood
as an “indispensible prerequisite and actual raison d’être” for a
work of art.79 As both art and architecture, the Mactar laby-
rinth awaits the viewer’s activation and animation because its
completion relies on a beholding subject to take the stage it
provides. The myth is no longer merely a prototype for
human experience; it is now a call to specific action. The
floor mosaic literally sets the stage for the viewer to occupy.
If the permanent exchange of different levels of percep-
tion is central to the function of many floor mosaics, it is
essential here. The labyrinth culminates at the doorstep of
the next room—the palaestra—rather than within the center 15 Threshold of the palaestra, with the culmination of the
labyrinth mosaic, Great East Baths, Mactar, Tunisia (artwork in
of the room it adorns. There is nowhere else to go from this the public domain; photograph by the author)
central threshold. The location of the labyrinth mosaic can-
not be accidental: the invitation for the viewer to displace his
own motion into the realm of myth is very clear. The laby- fostered by the ongoing spectacle in adjacent spaces. This
rinth mosaic lacks even a black border on the edge to distin- monumental mosaic is not a static given; we have no image of
guish it from neighboring spaces, and this lack of visual an event, and thus it is not possible to arrest the time of a
resolution prompts the viewer to participate in the illusion recorded narrative. Rather, it is a precarious, moving image,
 298 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
Downloaded by [University of West Florida] at 11:37 28 December 2014
whose complete lack of borders encourages the blurring
between the mythical realm of the floor and the activities
taking place in reality.80
Gilbert-Charles Picard believes that matches held in the
great palaestra of Mactar were organized by notables and
fought by professionals.81 Sand in the pit would have pro-
vided a smooth, soft surface on which athletes could be
thrown in wrestling matches and not be injured when they hit
the ground. This wrestling pit, surrounded by a U-shaped
colonnade, is positioned as if at the center of the semicircular
labyrinth of the exedra (Fig. 13). The labyrinth mosaic is
itself a symbol of a challenge, and its presence in the baths
would have lifted wrestling matches held there into the realm
of mythological reenactment or, at least, provided appropri-
ate and resonant mythological echoes.
A sculpted bas-relief nearly 6 feet 7 inches (2 meters) long
also contributed to the sporting atmosphere of the Mactar
baths (Fig. 16). It is not clear exactly from which part of the
baths this relief came, but Picard suspects it would have been
completed by another length of stone, and that the whole
may well have decorated the upper zone of a niche holding a
statue of a victorious athlete or donor.82 It displays six figures:
a chubby figure at right holding a wreath has been identified
as Eros;83 Hercules stands nearby with his ankles casually
crossed and his club at his feet. Also visible is the skin of the
Nemean lion, a beast strangled by means of a wrestling hold.
Hercules’ heavy labors, including his defeat of Antaeus in
wrestling, made him as natural a subject for the decor of
Roman baths as Theseus. Next to the hero are two pairs of
naked wrestlers. The first pair, one man shown flying through
the air on the way to a tough fall, illustrates a dramatic throw.
The arms and legs of the second pair are locked in combat:
each man tries to trip the other and get him in a headlock at
the same time. This relief easily conjoins human and divine
16 Bas-relief of wrestlers, both human
and divine, Great East Baths, Mactar,
Tunisia (artwork in the public domain;
photograph by the author)
athletics onto the same plane; the labyrinth mosaic should be
classed as another of these links.
The labyrinth, location of an epic wrestling match involv-
ing the hero credited with laying down the rules of the sport,
was just as fitting an adornment for a bath as a relief of
Hercules, god of the palaestra. At the Mactar baths, architec-
ture and decorative scheme collude to implicate the be-
holder in their surrounding narrative. It is not a static image
of victory but, rather, a prompt toward participation that uses
the mythic landscape underfoot and the beholder’s move-
ment to activate an open doorway and the space beyond.
Wrestling with Myth
The labyrinth is an especially appropriate subject for a floor
mosaic since it represents a journey that must itself be navi-
gated in order to be seen and experienced. This can be a
traversal shared between two realms: the viewer moves both
across a room and into a myth. In the baths of Roman Africa,
as bathers walked across mosaic images and patterns, the
pavements’ epic images and complex designs became the
stage for the viewer’s own actions. The pattern of the laby-
rinth implies movement, and it is essentially the construction
of a spatial and temporal program. It scales up the architec-
tural space it adorns by turning the floor surface into a vast
plane of heroic enterprise, and the labyrinth mosaics play
with the notion that the beholder might be transformed into
a second hero merely by making the journey to the laby-
rinth’s center. Labyrinth mosaics invite the beholder’s move-
ment to their centers, where the Minotaur awaits, whether in
the form of an image or in the form of a human opponent in
an impending wrestling match.
The depictions of Theseus and the Minotaur borrow much
from the lived experience of the baths, and specifically their
sporting ambiance. The poses and body types of the hero,
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION
famous for his wrestling prowess, link him with living wres-
tlers and pancratiasts, whether bulky or ideal, and with por-
traits of athletes found in many Roman baths. The tangles of
wrestling, the required twists and turns of the body, were
compared by Catullus to the structure of the labyrinth itself,
with the winding thread of Ariadne guiding Theseus from the
“inextricable entanglement of the building.”84 So, too, when
the early Christian moralist Tertullian speaks of wrestling, he
evokes the “binding twist” of the body and the “suppleness
that eludes.”85 The “many miles of walks or rides” that Pliny
the Elder describes as compressed into tessellated floors of
the labyrinth86 may also prompt a nexus between myth and
current life, harking back to the leisurely strolls taken by
bathers along the porticoes and through the long halls of the
larger bath complexes, which often featured statues of Her-
cules, athletes, and even bulls.87
There can be little doubt that context makes meaning for
the labyrinth mosaics, and vice versa: the labyrinths of the
Downloaded by [University of West Florida] at 11:37 28 December 2014
North African baths offered many prompts to blur athletic
and heroic activity, sometimes, as at Mactar, even relying on
the architecture of the space to provide doorways at once
both actual and mythological. Labyrinth mosaics are repre-
sentations of a spatial experience on whose surfaces the
otherwise unremarkable action of walking cannot be taken
for granted. In the realm of the Roman bath in North Africa,
their traversal easily takes on metaphoric meaning, as mosaics
energize lived spaces and the beholder’s actual path and the
horizon of myth are fused.
Rebecca Molholt received her PhD in art history from Columbia
University in 2008. Her dissertation on Roman floor mosaics was
completed during a fellowship at the Center for Advanced Study in
the Visual Arts at the National Gallery of Art, in Washington,
D.C. She is an assistant professor at Brown University [History of
Art and Architecture, Brown University, Providence, R.I. 02912,
rebecca_molholt@brown.edu].
Notes
I would like to thank the Center for Advanced Study in the Visual Arts at the
National Gallery of Art, Washington, D.C., and the American Academy in
Rome for funding my work on this project, and audiences at these institutions
for their thoughtful comments on earlier versions of these arguments. I am
very grateful to both institutions for funding years of fellowship in Rome and
in Washington, D.C., where these ideas gestated. Christine Kondoleon intro-
duced me to mosaics, and she and Bettina Bergmann have encouraged me
(and traveled with me) since the beginning. At Columbia, I was lucky enough
to be in Richard Brilliant’s first-ever mosaics seminar, and Francesco de
Angelis was always generous with his time and ideas. Roger Hanoune and
Fikret Yegül advised me on earlier versions of this article, first presented at the
Association Internationale pour l’Étude de la Mosaı̈que Antique in Portugal.
Additional thanks to Prof. Dr. Yegül for permission to publish a version of his
plan of the Mactar baths with the labyrinth mosaic inserted. Jean-Pierre
Darmon and Katherine Dunbabin kindly provided advice for seeking out
illustrations, and Lindsay Elgin generously assisted with the images. Heartfelt
thanks to many friends and colleagues who have offered critique and conver-
sation, including Susan Alcock, Sean Anderson, Michelle Berenfeld, John
Bodel, Sheila Bonde, John Cherry, James Frakes, Barbara Kellum, Dian Kriz,
Aı̈cha Malek, Elizabeth Marlowe, Douglas Nickel, James Trilling, and Hervé
Vanel. Many thanks to The Art Bulletin editor Karen Lang and the two
anonymous reviewers at The Art Bulletin for their many helpful and perceptive
comments on the manuscript and to Fronia W. Simpson for her copyediting.
This study has benefited most significantly from years of conversations with
Natalie Kampen, gift to us all.
1. Walter Benjamin to Gershom Scholem, October 22, 1917, quoted in
Michael Jennings, Bridget Doherty, and Thomas Levin, eds., The Work
299
of Art in the Age of Its Technological Reproducibility, and Other Writings on
Media (Cambridge, Mass.: Harvard University Press, 2008), 197.
2. Scholarly uneasiness with floor mosaics is surprising and long-lived: in
1991 Roger Ling described floor mosaics as “disturbing” and “uncom-
fortable” for modern viewers; Ling, Roman Painting (New York: Cam-
bridge University Press, 1991), 19 –20. Seeking to furnish evidence for
lost monumental paintings, scholars have long privileged figural mosa-
ics; the most famous Roman floor mosaics are the so-called Bildmosai-
ken, “paintings in stone.” See, for example, Bernard Andreae, Antike
Bildmosaiken (Mainz: Philipp von Zabern, 2003); and Ada Cohen, The
Alexander Mosaic: Stories of Victory and Defeat (Cambridge: Cambridge
University Press, 1997). Scholars who consider the narrative possibili-
ties inherent in horizontal display and who consider nonfigurative mo-
saics have not assumed mosaics were created in imitation of paintings.
See esp. John Clarke, Roman Black-and-White Figural Mosaics (New York:
New York University Press, for the College Art Association of America,
1979); Christine Kondoleon, Domestic and Divine: Roman Mosaics in the
House of Dionysos (Ithaca, N.Y.: Cornell University Press, 1995); Ellen
Swift, Style and Function in Roman Decoration: Living with Objects and Inte-
riors (Burlington, Vt.: Ashgate, 2009); and Norman Bryson, Looking at
the Overlooked: Four Essays on Still Life Painting (Cambridge, Mass.: Har-
vard University Press, 1990). Bettina Bergmann considers the “inhabit-
ant’s supposed location in the room” when viewing circus mosaics such
as the one from Piazza Armerina, Sicily, and another from a villa at
Silin, Libya; Bergmann, “Pictorial Narratives of the Roman Circus,” in
Le cirque et son image, ed. J.-M. Roddaz and J. Nélis-Clement (Bordeaux:
Ausonius, 2009), esp. 371–76. In her monumental survey of Greek and
Roman mosaics, Katherine M. D. Dunbabin notes that mosaics, given
their durability, “offer an invaluable contribution to our knowl-
edge . . . of major painting.” She is, however, quick to point out that
mosaics are a significant art form in their own right, and it is hoped
her book will render such statements unnecessary for future publica-
tions; Dunbabin, Mosaics of the Greek and Roman World (Cambridge:
Cambridge University Press, 1999), 1.
3. Stephen Cosh and David Neal, “Roman Mosaics,” Current Archaeology
157 (May 1998): 18. These authors recommend that mosaics are best
imaged by means of paintings, and indeed, painstaking watercolors of
floor mosaics accompany their text.
4. Here I am drawing on the work of Eugene Y. Wang, especially his fasci-
nating chapter “Watching the Steps: Peripatetic Viewing in Medieval
China,” in Visuality before and beyond the Renaissance: Seeing as Others Saw,
ed. Robert S. Nelson (Cambridge: Cambridge University Press, 2000),
116 –38.
5. The contextual deployment of myth and the resulting interactions be-
tween life and myth have been especially well explored by Paul Zanker
and Björn Ewald, Mit Mythen Leben: Die Bilderwelt der römischen
Sarkophage (Munich: Hirmer, 2004). The authors trace the selection of
particular moments chosen from grand narrative schemes and the ar-
tistic display of these moments in specific contexts to prompt meaning-
ful and powerful links with situations in daily life, up to and including
the direct assimilation of beholders with mythological protagonists.
6. Michèle Blanchard-Lemée et al., Mosaics of Roman Africa: Floor Mosaics
from Tunisia (New York: George Braziller, 1996), 190. See also nn. 40,
41 below.
7. In 1977 Wiktor A. Daszewski collected and published sixty-two Roman
mosaics depicting themes of the labyrinth, from across the Roman
world; Daszewski, La mosaı̈que de Thésée: Études sur les mosaı̈ques avec
représentations du labyrinthe, de Thésée et du Minotaure (Warsaw: PWN, Édi-
tions Scientifiques de Pologne, 1977). To these we can add the laby-
rinths from the Baths of Julia Memmia at Bulla Regia (Tunisia) and a
small threshold labyrinth from a Roman house in second–third century
CE Conimbriga (Portugal), not featured in Daszewski’s catalog, for a
total of sixty-four labyrinth mosaics. The Baths of Julia Memmia have
been most recently published by Roger Hanoune, “Décor du monu-
ment: Les pavements mosaı̈ques,” in Recherches archéologiques franco-
tunisiennes à Bulla Regia, vol. 2, pt. 1, Les thermes memmiens: Étude architec-
turale et histoire urbaine, by Henri Broise and Yvon Thébert (Rome:
École Française de Rome, 1993), 245–71. The Conimbriga threshold
labyrinth is cat. no. 129 in Hermann Kern, Through the Labyrinth: De-
signs and Meanings over 5,000 Years (Munich: Prestel, 2000), 91, with
earlier bibliography. It should be noted that Daszewski’s catalog in-
cludes eight mosaics, probably all from houses, that show the combat
between Theseus and the Minotaur without any enframing labyrinth
surround. These eight vignettes almost all show the interactions be-
tween hero and monster taking place outside the labyrinth, with the
heavy walls and doors of the structure serving as a backdrop to the
scene. I have omitted these from my calculations because they lack an
associated labyrinth surround, hence my total of fifty-six.
8. Daszewski divided his study by typology; I divide the corpus by context
and location. As Daszewski notes (La mosaı̈que de Thésée, 99), the pave-
ments are almost exclusively from the western provinces of the empire,
and, despite the abundance of mosaics surviving from Roman prov-
 300 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
inces in Syria, Egypt, Anatolia, Israel, and Palestine, there seem to be
no labyrinth mosaics from any site farther east than Cyprus. Twenty-
eight labyrinth mosaics come from houses and villas. About half the
labyrinth mosaics of North Africa (seven of sixteen) come from baths.
This is a striking preponderance of a theme already notably popular in
North Africa. None of the seven labyrinth mosaics in Roman Iberia
comes from baths, and only two of the seventeen surviving Italian laby-
rinth mosaics come from baths.
9. Herodotus, 2.48; Pliny, Natural History 36.86.90; Plutarch, Theseus esp.
15, 19, 21; Pausanias, 1.27.10; Apollodorus, Bibliotheca 3.1.4 and Epitome,
1.7–10; and Strabo 8.6.2, among others.
10. Broise and Thébert, Recherches, 41: “This subject was often represented
and seems to have been especially prized in the salles de répos of the
baths, where it constituted perhaps a sort of visual game” (my transla-
tion).
11. Daszewski, La mosaı̈que de Thésée, 95–96; and Katherine Dunbabin, “Ba-
iarum Grata Voluptas: Pleasures and Dangers of the Baths,” Papers of the
British School at Rome 57 (1989): 40. See also n. 25 below, on the apo-
tropaic resonance of mosaics.
12. Daszewski, La mosaı̈que de Thésée, 95.
13. Ibid., 99.
14. Federica Cordano, “Il labirinto come simbolo grafico della città,”
Mélanges de l’École Française de Rome: Antiquité 92, no. 1 (1980): 7–15;
Downloaded by [University of West Florida] at 11:37 28 December 2014
Gianna Dareggi, “I mosaici con raffigurazione del labirinto: Una varia-
zione sul tema del ’centro,’ ” Mélanges de l’École Française de Rome: Anti-
quité 104, no. 1 (1992): 281–92.
15. Most recently in Yvon Thébert, Thermes romaines d’Afrique du nord et leur
contexte méditerranéen: Études d’histoire et d’archéologie, Bibliothèque des
Écoles Françaises d’Athènes et de Rome, fasc. 315 (Rome: École Fran-
çaise de Rome, 2003), 479 n. 110. Also Dunbabin, Mosaics of the Greek
and Roman World, 11. See also Simone Wiedler, “Labyrinthdarstel-
lungen,” in Aspekte der Mosaikausstattung in Bädern und Thermen des
Maghreb (Hamburg: Dr. Kovac, 1999), 55–57.
16. I discuss only the four best-preserved examples of North African bath
labyrinths here: Hippo Regius in Algeria, and Belalis Maior, Thuburbo
Maius, and Mactar, all in Tunisia. The other three bath labyrinth mosa-
ics come from Dellys and Rusguniae (Algeria), and the Baths of Julia
Memmia (Tunisia). To make a grand total of sixteen, the other nine
North African labyrinths come from various settings: six from houses
or villas in Libya, Algeria, and Tunisia (and none of these six defini-
tively from bath suites), two from churches at sites in modern-day Alge-
ria, and one from a tomb (Hadrumentum, Tunisia). Numbers are
based on Daszewski, La mosaı̈que de Thésée; and Hanoune, “Décor du
monument.” Hédi Slim, “La mosaı̈que du labyrinthe de Thysdrus,” An-
tiquités Africaines 15 (1980): 207– 8, suggests that only six of the sixteen
North African labyrinths are from baths, but he considers the
Thuburbo labyrinth (Figs. 4 and 5) to be from a villa. This may be a
domestic mosaic, but it is certainly from a bath suite.
17. See David Lindberg, Theories of Vision from Al’Kindi to Kepler (Chicago:
University of Chicago Press, 1976), 1–17; David Summers, Vision, Reflec-
tion, and Desire in Western Painting (Chapel Hill: University of North
Carolina Press, 2007); A. Mark Smith, Ptolemy’s Theory of Visual Perception
(Philadelphia: American Philosophical Society, 1996); Albert Lejeune,
Euclide et Ptolémée: Deux stades de l’optique géométrique grecque (Louvain:
Bibliothèque de l’Université, 1948); and idem, L’optique de Claude
Ptolémée, dans la version latin d’après l’arabe de l’émir Eugène de Sicile (New
York: E. J. Brill, 1989). Henri Lavagne briefly touched on the problem
of mosaics and the optic versus the haptic in La mosaı̈que: Que sais-je?
(Paris: Presses Universitaires de France, 1987), 11–12. Optical and hap-
tic experiential reactions have been analyzed by architectural histori-
ans; see esp. Diane Favro’s re-created urban walks in The Urban Image of
Augustan Rome (New York: Cambridge University Press, 1996).
18. Wall paintings at the Basilica of Herculaneum and the Houses of Mar-
cus Gavius Rufus and Marcus Lucretius Fronto at Pompeii show only
the labyrinth’s deep entrance. In each case, the hero is present at the
doorway, his journey already accomplished, and his defeat of the Mino-
taur evidenced by the hero’s relaxed stance and/or the vanquished
body of the beast.
19. Clarke, in Roman Black-and-White Figural Mosaics, opened a discussion of
the viewer’s relationship to the mosaic floor underfoot. He presented
the concept of “kinaesthetic address,” which he defined as “the use of
the figure to influence spectator movement” (21). Figures underfoot,
he suggests, might prompt the walking viewer to go to the left, turn to
the right, or continue through a nearby door, according to what he
calls a “traffic-flow suggestion” (33). These concepts have been further
elaborated by Ellen Swift, who studied geometric pavements, especially
thresholds, and their use as apotropaic symbols and identifying mark-
ers for different rooms. See “Interiors: Non-figurative Floor Mosaics
and Other Domestic Decoration,” in Style and Function, 27–104.
20. Maurice Merleau-Ponty, “Cézanne’s Doubt,” in The Merleau-Ponty Aesthet-
ics Reader: Philosophy and Painting, ed. Galen A. Johnson (Chicago:
Northwestern University Press, 1993), 8.
21. Anthologia Latina 1.208.
22. While discussing the Hunting Baths of Leptis Magna, Fikret Yegül sug-
gests that the dramatic scenes of danger at the hunt “might have re-
called, in a general and distant way, the literary and heroic idea of
bathing as a pleasurable reward for those who have endured physical
hardship”; Yegül, Baths and Bathing in Classical Antiquity (Cambridge,
Mass.: MIT Press, 1992), 184.
23. Indeed, Neptune is at the beginning and the end of the story. Apol-
lodorus tells how Minos prayed to Poseidon for a bull to be sent up
from the depths of the sea, but when the god answered his prayers, the
king failed to sacrifice the bull to Poseidon, as he had promised (The
Library 3.1.4). The eventual sacrifice of the Minotaur, fruit of an unnat-
ural union, might have been pleasing to Poseidon, and perhaps the
story of the labyrinth and the death of the Minotaur might also have
been appropriate for bath decoration because of this link with water
and the god of the sea.
24. Dunbabin, “Baiarum Grata Voluptas,” 40.
25. Images and patterns in other types of settings, such as thresholds, can
also be understood to have apotropaic resonance. See Swift, “Interiors.”
The idea of the labyrinth as a puzzle for the eyes, trapping the gaze,
lent it considerable apotropaic power in the Roman world. Small laby-
rinths and knots, impossible to untangle, occasionally appear at thresh-
olds, where, it was hoped, they could bind envy and ill will and prevent
these forces from entering the house. On knots, see Eunice Maguire,
Henry Maguire, and Maggie Duncan-Flowers, Art and Holy Powers in the
Early Christian House (Urbana: University of Illinois Press, 1989), 3– 4.
On the “binding” language of spells, see Christopher Faraone, “The
Agonistic Context of Early Greek Binding Spells,” in Magika Hiera, ed.
Faraone and Dirk Obbink (New York: Oxford University Press, 1991),
3–32; and Katherine Dunbabin and Matthew W. Dickie, “Invidia rump-
antur pectora: The Iconography of Phthonos/Invidia in Graeco-Roman
Art,” Jahrbuch für Antike und Christentum 26 (1983): 7–37. See also Pliny,
Natural History 36.19.85, where he is careful to note that “We must not,
comparing this last to what we see delineated on our mosaic pave-
ments, or to the mazes formed in the fields for the amusement of chil-
dren, suppose it to be a narrow promenade along which we may walk
for many miles together; but we must picture to ourselves a building
filled with numerous doors, and galleries which continually mislead the
visitor, bringing him back, after all his wanderings, to the spot from
which he first set out.” Pliny, Natural History, trans. D. E. Eichholz
(Cambridge, Mass.: Harvard University Press, 1962), 340. It should be
noted that the labyrinth mosaics are unicursal, offering only a single
route forward, and thus, even if the single path might be difficult to
follow with the eyes, their lack of “puzzle” would be easily apparent.
Nonetheless, in each case, the visual scintillation of the pattern can
impede optical clarity. See also n. 31 below, on mosaics of Medusa.
26. One parallel for this study can be found in the work of Timothy
O’Sullivan, who examines the Odyssey Landscapes (in the Vatican Mu-
seums) with attention to the painted portico that surrounds the fres-
coes, allowing the beholder to “walk with Odysseus,” both physically
and mentally, along the course of the hero’s journeys; O’Sullivan,
“Walking with Odysseus: The Portico Frame of the Odyssey Land-
scapes,” American Journal of Philology 128 (2007): 497–532.
27. The Baths of Theseus and the Minotaur at Belalis Maior (Tunisia) were
excavated in 1960 – 65 and are named after this mosaic. The baths have
been published in Thébert, Thermes romaines, 132–33, with a schematic
plan at pl. 33, 4. See also Ammar Mahjoubi, “Le thème du Labyrinthe et
du Minotaure figuré sur une mosaı̈que de Belalis Major (Henchir
el-Faouar),” Africa (1969 –70, published 1972): 335– 40; and idem, Recher-
ches d’histoire et d’archéologie à Henchir-El-Faouar (Tunisie) (Tunis: Publications
de l’Université de Tunis, 1978), 209 –27. According to Daszewski, La
mosaı̈que de Thésée, 123, the mosaic is currently in a local storage house
near the site. The labyrinth mosaic also features in Kern, Through the Laby-
rinth, 93, cat. no. 140. Marta Novello refers to the mosaic as adorning the
“House of Theseus and the Minotaur,” at Belalis Maior, where only the
baths survive. See Novello, Scelte tematiche e committenza nelle abitazioni
dell’Africa proconsolare: I mosaic figurati (Pisa: Fabrizio Serra, 2007), 225.
28. See Mahjoubi, “Le thème du Labyrinthe,” 340; and Daszewski, La
mosaı̈que de Thésée, 52, pl. 18.
29. A floor mosaic of Theseus and the Minotaur from a Roman house in
Gurgi, Libya (ca. 200) does not feature a labyrinth plan but shows a
scene of the hero dragging the monster out the door of the labyrinth,
under the watchful gaze of Ariadne. An adjacent vignette shows a still
life of rabbit, chicken, and fruit. See Daszewski, La mosaı̈que de Thésée,
119 –20, no. 45, pl. 36; and Kern, Through the Labyrinth, 101.
30. For one example, see the late third-century mosaic from the baths at
Thina, Tunisia, where at least twenty marine narratives are collaged
onto a single watery surface. See Nabiha Jeddi, “Étude descriptive et
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION
analytique des mosaı̈ques de Thaenae (Thina en Tunisie)” (PhD diss.,
Université de Paris–IV Sorbonne, 1990); J. Thirion, “Un ensemble ther-
mal avec mosaı̈ques à Thina (Tunisie),” Mélanges de l’École Française de
Rome: Antiquité 69 (1957): 207– 46; and Katherine M. D. Dunbabin, Mo-
saics of Roman North Africa: Studies in Iconography and Patronage (Oxford:
Clarendon Press, 1978), 273.
31. As with the many mosaics of Medusa, also a frequent feature in Roman
baths, the nearly abstract pattern of most of the floor is also doing nar-
rative work: manifesting the terrible power of her vision in a kaleido-
scopic swirl of pattern around her head. See, for example, a second-
century mosaic of Medusa from a tepidarium in Dar Zmela, now in the
Sousse Museum, Tunisia. See also Louis Foucher, Inventaire des
mosaı̈ques: Sousse (Tunis: Institut Nationale d’Archéologie et Arts, 1960),
121–22, no. 57.274, pl. 67; Dunbabin, Mosaics of Roman North Africa,
163 n. 149 and 271 n. 30b; and Carolyn McKeon, “The Iconography of
the Gorgon Medusa in Roman Mosaic” (PhD diss., University of Michi-
gan, 1983), 291–93, no. 64. Suzanne Germain, Les mosaı̈ques de Timgad
(Paris: Éditions du Centre National de la Recherche Scientifique,
1969), 90, notes that the design of such mosaics evokes Medusa’s aegis.
32. Natural History 36.19.85, trans. Eichholz, 67. Pliny is writing about a lab-
yrinth in Egypt, which he declares to be still extant, and which he as-
serts must have served as the model for Daedalus’s labyrinth in Crete.
Pliny goes on to mention the many miles that are compressed by laby-
rinth mosaic floors and by the maze games played by boys running
Downloaded by [University of West Florida] at 11:37 28 December 2014
around in the Campus Martius.
33. See John Kraft, “The Cretan Labyrinth and the Walls of Troy: An Anal-
ysis of Roman Labyrinth Designs,” Opuscula Romana 15, no. 6 (1985):
79 – 86; Anthony Phillips, “The Topology of Roman Mosaic Mazes,”
Leonardo 25, nos. 3– 4 (1992): 321–29; and Penelope Doob, The Idea of
the Labyrinth from Classical Antiquity through the Middle Ages (Ithaca, N.Y.:
Cornell University Press, 1990).
34. The sensitive study of Roger Hanoune on the Baths of Julia Memmia is
an exception; Hanoune, “Décor du monument.”
35. Pausanias 1.39.3, as cited in E. Norman Gardiner, “Wrestling,” Journal
of Hellenic Studies 25 (1905): 19. See also idem, “Wrestling (Contin-
ued),” Journal of Hellenic Studies 25 (1905): 263–93. Further discussion
of this tradition can be found in Michael Poliakoff, Combat Sports in the
Ancient World: Competition, Violence, and Culture (New Haven: Yale Uni-
versity Press, 1987), 46, 136. One could also compare these works with
Roland Barthes’s essay on wrestling, “The World of Wrestling,” in My-
thologies (New York: Hill and Wang, 1972), 15–25.
36. Daszewski, La mosaı̈que de Thésée, 94. In her magnificent article on bath
decor from 1989, Dunbabin (“Baiarum Grata Voluptas,” 24) explored
the “world of beauty and luxury which lay at the heart of the bath aes-
thetic,” but she too knowingly left aside the sporting aspect of the
baths.
37. Yegül (Baths and Bathing, 185, with sources) points out that these
spaces were suitable not only for games and gymnastic performances
but also for banquets. An honorific inscription from Africa praises a
man who “gave a banquet to the entire population and a gymnastic
contest; at the same spectacle he also showed boxers . . . ,” trans. Anne
Mahoney, Roman Sports and Spectacles: A Sourcebook (Newburyport, Mass.:
Focus Publishing, 2001), 64. As noted by Mahoney, the date and the
name of the honoree are unknown.
38. See also Thébert, Thermes romaines, 67– 68. For athletic scenes from
baths in Italy, and suggestions of their appropriateness to the location,
see Zahra Newby, “Greek Athletics as Roman Spectacle: The Mosaics
from Ostia and Rome,” Papers of the British School at Rome 70 (2002):
177–203, where she suggests that mosaic figures shown beneath living
bathers “could serve as a model to which to aspire” (200).
39. According to Garth Fagan, Roman (men) embarked on the bathing
process only “after a good sweat had been worked up”; Fagan, Bathing
in Public in the Roman World (Ann Arbor: University of Michigan Press,
1999), 10. He is referring only to Roman men in this discussion. How
Roman women, families, or children made use of baths, both public
and private, has not yet been a topic of scholarly study.
40. See Newby, “Greek Athletics as Roman Spectacle,” 180, with citations.
Also Inge Nielsen, Thermae et Balnea: The Architecture and Cultural History
of Roman Public Baths (Aarhus: Aarhus University Press, 1990), 144.
41. Mustapha Khanoussi, “Pugilist Spectacles and Athletic Games in Pro-
consular Africa,” in Stories in Stone: Conserving Mosaics of Roman Africa,
ed. Aı̈cha Ben Abed (Los Angeles: J. Paul Getty Museum Publications,
2006), 79 –91, including discussion of mosaics of wrestlers and various
athletic games from the early fourth-century baths at Gafsa (now in the
Gafsa Archaeological Museum, Tunisia) and from the third-century
Baths at Gigthis, now in the Bardo Museum, Tunis. See also Mustapha
Khanoussi, “Les spectacles de jeux athlétiques et de pugilat dans
l’Afrique romaine,” Mitteilungen des Deutschen Archäologischen Instituts:
Römische Abteilung 98 (1991): 315–22; and idem, “Spectaculum pugilium
et gymnasium: Compte rendu d’un spectacle de jeux athlétiques et de
301
pugilat, figuré sur une mosaı̈que de la région de Gafsa (Tunisia),”
Académie des Inscriptions et Belles-Lettres, Comptes Rendus, 1988: 543– 60.
For a recent compendium, see Simone Wiedler, “Gymnische Agone
oder Certamina Graeca,” in Wiedler, Aspekte der Mosaikausstattung, 57–72.
42. Joyce M. Reynolds and John B. Ward Perkins, eds., Inscriptions of Roman
Tripolitania (London: British School at Rome, 1952), no. 103a, cited in
Fagan, Bathing in Public, app., 243, no. 35.
43. The Baths of the Labyrinth at Thuburbo Maius have been published
most recently in Thébert, Thermes romaines, 172–73. Mosaics are pub-
lished in Margaret Alexander, Aı̈cha Ben Abed-Ben Khader, S. Besrour-
Ben Mansour, David Soren, Corpus des mosaı̈ques de Tunisie, vol. 2, fasc.
1, Thuburbo Maius: Les mosaı̈ques de la région du Forum (Tunis: Institut
National d’Archéologie et d’Arts, 1980), 20 –31, as end 3rd– beginning
4th century. Thébert (ibid.) concurs, and Dunbabin, Mosaics of Roman
North Africa, 274, says “?4th century.” For the architecture, see Louis
Drappier, “Les thermes de Thuburbo Maius,” Bulletin Archéologique du
Comité des Travaux Historiques et Scientifiques, 1920: 55–75. Boxers face
off as well in a threshold mosaic at the House of the Boxers in Utica,
discussed below. See Margaret Alexander, S. Besrour, Mongi Ennaı̈fer
et al., Corpus des Mosaı̈ques de Tunisie, vol. 1, fasc. 3, Utique, mosaı̈ques
sans localisation précise et El Alia (Tunis: Institut National d’Archéologie
et d’Arts, 1976), 276, cat. no. 246, pl. 2. Boxers feature in mosaics of
the Thermae of the Pugilists at Thina, as well as in the second-century
mosaics of the Baths of Massongex in the Valais, Switzerland. See Jean-
Paul Thuillier, Le sport dans la Rome antique (Paris: Éditions Errance,
1996), 141. The athlete portraits in the mosaics of the Baths of Cara-
calla at Rome are another obvious parallel (see n. 67 below), as are the
mosaic portraits of athletes from the Baths of Porta Marina, Ostia. Re-
garding a black-and-white mosaic from Ostia, boxing, wrestling, and
associated literary sources, see Christopher Jones, “The Pancratiasts He-
lix and Alexander on an Ostian Mosaic,” Journal of Roman Archaeology
11 (1998): 293–98. For general surveys, see Zahra Newby, “The Athletic
Ideal in the Second Sophistic,” a chapter in her dissertation, “Educated
Fantasies: Interpreting the Visual Arts in the Second Sophistic” (PhD
diss., Courtauld Institute of Art, 2000); and idem, “Greek Athletics as
Roman Spectacle,” 177–203; Poliakoff, Combat Sports; and Jean-Paul
Thuillier, “Athletic Exercises in Ancient Rome,” European Review 12
(2004): 415–26. See also n. 35 above, on wrestling.
44. Thébert, Thermes romaines, 455, citing several third-century inscriptions,
and see also his n. 55.
45. Ibid., 172. The smaller North African baths generally did not have pa-
laestrae. See Yegül, Baths and Bathing, 186 – 87. This is the labyrinth
mosaic Slim classed as from a villa, rather than a bath; see n. 16 above.
46. The narrow threshold between the frigidarium and tepidarium was
marked by a mosaic of sandals. See plan in Thébert, Thermes romaines,
pl. LVII, 3; and Alexander et al., Thuburbo Maius, 27 and pl. IX. Such a
mosaic of footwear at a threshold certainly addresses a walking viewer,
indicating that the temperature underfoot was about to change dramat-
ically. At this point, precautionary measures would be taken to protect
unshod feet from heated floors, or, conversely, bath sandals no longer
needed could be taken off.
47. Natural History 36.19.84 – 86, trans. Eichholz, 69.
48. Such a perspectival illusion has been noted by Jean Lassus, discussing a
mosaic from the markets at Hippo Regius; Lassus, “La mosaı̈que ro-
maine: Organisation des surfaces,” in La mosaı̈que Gréco-Romaine II, ed.
Henri Stern and Marcel Le Glay (Paris: A. & J. Picard, 1975), 337 and
pl. CLIX, 1.
49. Apollodorus, The Library 3.1.4, trans. J. Frazer (Cambridge, Mass.: Har-
vard University Press, 1939).
50. Poliakoff, Combat Sports, 8. See also Gardiner, “Wrestling,” 24. For more
on the pankration, see n. 58 below.
51. Dio Chrysostom, 8.26; and Galen, Exhortation for Medicine 9 –14; dis-
cussed in Stephen G. Miller, Arete: Ancient Writers, Papyri and Greek In-
scriptions on the History and Ideals of Greek Athletics and Games (Berkeley:
University of California Press, 1991). According to Galen, Roman wres-
tlers seeking to add strategic heft were said to gorge themselves on
flesh and blood (Exhortation for Medicine 9 –14). This is also, however, a
reasonably apt description of the Minotaur’s diet. He is shown in some
labyrinth mosaics (as at Thuburbo Maius) surrounded by the detritus
of past victims, his own meals of flesh and blood.
52. Wulf Raeck has studied the late-antique trend of actualizing mythologi-
cal scenes by using contemporary clothing and hairstyles for mythologi-
cal characters in his book, Modernisierte Mythen: Zum Umgang der Spätan-
tike mit klassischen Bildthemen (Stuttgart: Steiner, 1992). At Thuburbo
Maius, Theseus’s bulla (amulet), his short hairstyle, the chunky body
types of both figures, and the recognizable wrestling holds offer an up-
dated version of the labyrinth story. The body type of Theseus at Bela-
lis Maior (seen in Fig. 2) is decidedly more heroic. The nonheroic
body type is unusual for mosaics of Theseus—I know of only one addi-
tional example, from the Villa Domizia on the island of Giannutri, It-
 302 ART BULLETIN SEPTEMBER 2011 VOLUME XCIII NUMBER 3
aly. This pavement, entirely black-and-white and dated to 150 –200 CE,
shows the two protagonists kneeling in a narrow rectangular space at
the labyrinth’s center. Theseus is nude and quite thick around the
middle. Nonetheless, the battle is about to be won. With one hand,
Theseus has a good grip on one of the Minotaur’s horns, and the hero
again wields his pedum in his other hand. See Daszewski, La mosaı̈que de
Thésée, cat. no. 29, pl. 15.
53. See Gardiner, “Wrestling (Continued),” 265, for discussion of the Oxy-
rhynchus papyrus 3.466.
54. Plutarch, Theseus 19.3, trans. B. Perrin (Cambridge, Mass.: Harvard
University Press, 1948).
55. Apollodorus, Epitome 1.9; Plutarch, Theseus 17.3, trans. Perrin. See also
Roger Ling, “The Casa della Caccia Antica at Pompeii,” Journal of Ro-
man Archaeology 18 (2005): 597–98, for a discussion of Theseus’s fight-
ing accoutrements in Pompeian wall painting, with bibliography.
56. As referenced by Ovid, Heroides 10.101–2.
57. For in-depth discussion of the rules of wrestling, see Gardiner, “Wres-
tling,” 14 –31; Gardiner, “Wrestling (Continued),” 263–93; and Polia-
koff, Combat Sports.
58. Of the sixty-two mosaics featured in Daszewski’s catalog, La mosaı̈que de
Thésée, twenty-six show a scene of combat between Theseus and the Mi-
notaur. Fully twenty-two of these combat scenes show the Minotaur
brought to his knees. The kicking pose, however, is a signal that this
Downloaded by [University of West Florida] at 11:37 28 December 2014
match may be more of a pankration-style encounter, in which hitting
and kicking were allowed. Fighting continued on the ground in the
pankration, and the match ended only when one adversary acknowl-
edged defeat. In the myth and on the mosaics, this battle was, of
course, the end of the Minotaur, and, unlike wrestling, the pankration
could be a life-and-death matter. See Gardiner, “Wrestling,” 19 –21,
with sources on the pankration. See also Poliakoff, Combat Sports, 54.
Galen’s satire on professional athletes gave the prize to a donkey for
demonstrating such prowess in kicking (Protrepticus 13). Other mosaic
labyrinths intimate combat even if the two protagonists are not pre-
sented, showing, for example, two crossed lances and a shield at the
labyrinth center (a bath complex at Chusclan, France; see Daszewski,
cat. no. 15), or usurping the immediate architectural surround by plac-
ing the labyrinth’s center at the threshold of the bath’s palaestra (Mac-
tar, Tunisia, see below).
59. On wrestlers’ hairstyles, see Gardiner, “Wrestling,” 18, citing Philostra-
tus, Imagines 2.32, and Plutarch, Aratus 2.3.6, among others. A version
of this hairstyle can be seen in many mosaics of athlete portraits, such
as the bust-length mosaic portraits of stocky athletes wearing necklaces
featured in floor mosaics at the Baths of Thapsus, from ca. 300 CE. See
N. Ben Lazreg, “Byzacene cotière: Athlètes,” in Aı̈cha Ben Abed-Ben
Khader, Image de pierre: La Tunisie en mosaı̈que (Paris: Ars Latina, 2003),
492–93, 534 and figs. 312, 313.
60. On the Utica mosaic, see Alexander et al., Utique, 276, cat. no. 246, pl.
2. The mosaics of Piazza Armerina, including the bath attendants, are
most comprehensively published in Gino Gentili, La villa romana di Pi-
azza Armerina Palazzo Erculio, vol. 3, I mosaici figurati: Descrizione e interpre-
tazione (Osimo: Fondazione Don Carlo, 1999). See also Andrea Caran-
dini, Andreina Ricci, and Mariette de Vos, Filosofiana: The Villa of Piazza
Armerina; The Image of a Roman Aristocrat in the Time of Constantine (Pa-
lermo: S. F. Flaccovio, 1982).
61. Amulets appearing within mosaic images might echo the apotropaic
import of the labyrinth motif itself. On the apotropaic nature of the
labyrinth, see Daszewski, La mosaı̈que de Thésée, 95–96; Dunbabin, “Bai-
arum Grata Voluptas,” 40; and see also n. 25 above. Amulets (along with
bells and symbols) probably also served as apotropaic functionaries out-
side tombs. See Donatello Nuzzo, “Amulet and Grave in Late Antiquity:
Some Examples from Roman Cemeteries,” in Burial, Society and Context
in the Roman World, ed. John Pearce, Martin Millett, and Manuela
Struck (Oxford: Oxbow Books, 2000), 249 –55. Surviving examples (in
bronze) of amulets that parallel those worn by athletes and heroes on
North African mosaics can be seen in Stephanie Boucher, Bronzes ro-
mains figurés du Musée des beaux-arts de Lyon (Lyons: Éditions de Boc-
card, Paris, 1973), cat. nos. 311–17, all long pendants and largely phal-
lic in import. Cat. no. 322 in Boucher, a bronze caduceus 47⁄8 inches
(12.5 cm) long, looks to be the kind of amulet that could easily be re-
alized in rope or leather cord, a more friable form of necklace or amu-
let that does not survive for us today but might more closely approxi-
mate what the mosaic athletes often appear to be wearing. In the
Museum of Fine Arts, Boston, there is a pendant amulet of a bull’s
head, perhaps an interesting option for a wrestler (though this exam-
ple is dated to 600 BCE and thus much earlier than any mosaics dis-
cussed here). See Mary Comstock and Cornelius Vermeule, Greek, Etrus-
can and Roman Bronzes in the Museum of Fine Arts, Boston (Greenwich,
Conn.: New York Graphic Society, 1971), cat. no. 275.
62. Gardiner (“Wrestling” and “Wrestling [Continued]”) traces these wres-
tling poses and motifs back to early black-figure vases. See also discus-
sions in Poliakoff, Combat Sports, passim.
63. Erwan Marec, “Le thème du labyrinthe et du Minotaure dans la
mosaı̈que romaine,” in Hommages à Albert Grenier, vol. 3 (Brussels: Lato-
mus, Revue d’Études Latines, 1962), 1094 –112. Also idem, Hippone la
Royale: Antique Hippo Regius (Algiers: Imprimerie Officielle, 1954), 100;
and Wiedler, Aspekte der Mosaikausstattung, 236, with previous bibliogra-
phy.
64. Daszewski, La mosaı̈que de Thésée, 101.
65. Plutarch, Theseus 19.
66. Only a few mosaics show the thread traversing the path of the laby-
rinth. One Roman labyrinth mosaic, today in the Kunsthistorisches Mu-
seum, Vienna, shows a red thread winding through a black labyrinth
and ending in the central square zone where Theseus has the Mino-
taur already on his knees (Daszewski, La mosaı̈que de Thésée, cat. no. 7).
Another, from a fourth-century church in Algeria, shows the thread
curving along to begin the journey, and then petering out quite soon
(ibid., cat. no. 4). At the Roman villa at Nea Paphos, Cyprus, the laby-
rinth is traversed by a band of guilloche that entirely encircles the cen-
ter scene (this pavement is the focus of ibid. and his cat. no. 8). At the
Baths of Julia Memmia (Bulla Regia, Tunisia), the labyrinth’s path is
followed by a laurel garland; see Hanoune, “Décor du monument.” A
sixth-century floor in S. Vitale, Ravenna, offers a path in opus sectile to
the labyrinth’s blank center (Daszewski, La mosaı̈que de Thésée, cat. no.
38); this labyrinth is composed like the one at Mactar discussed below:
all thread, no walls, no hero, no Minotaur (ibid., cat. no. 57). In addi-
tion to the Mactar labyrinth, only three other labyrinths show the ball
of thread at the labyrinth’s center—all from North African baths—the
ones at Hippo Regius and Belalis Maior, discussed here, and another
from Dellys, Algeria (ibid., cat. no. 3), where the thread is pictured in
the central battle scene.
67. For the athlete mosaics from Thapsus, from ca. 300, see Ben Abed-Ben
Khader, Image de pierre, 492–93, 534. Other mosaics of athletes in North
Africa can be found at the Antonine Baths in Carthage and baths at
the following sites: Bou Arkoub, Gigthis, Thina, Utica, Tébessa, Cher-
chel, Baten Zammour. See Khanoussi, “Pugilist Spectacles,” “Les specta-
cles,” and “Spectaculum pugilium.” For the athlete mosaics in the pa-
laestra exedrae of the Baths of Caracalla in Rome, see Antonio
Insalaco, “I mosaici degli atleti dalle Terme di Caracalla: Una nuova
indagine,” Archeologia Classica 41 (1989): 283–327; and Giuseppina Ghi-
rardini, “Die im Jahre 1824 bei den Grabungen Egidio Girolamo Di
Velos in den Caracallathermen aufgedeckten Athletendarstellungen,”
in Die Sammlung antiker Mosaiken in den Vatikanischen Museen by Klaus
Werner (Vatican City: Monumenti Musei e Gallerie Pontificie, 1998),
217–51.
68. Wolfgang Iser, Prospecting: From Reader Response to Literary Anthropology
(Baltimore: Johns Hopkins University Press, 1989), esp. the chap. “In-
teraction between Text and Reader,” 31– 41.
69. A similar bust appears at the center of a labyrinth mosaic from a Ro-
man baths at Stolac (Daszewski, La mosaı̈que de Thésée, cat. no. 29), and
the Minotaur appears alone in the center of two mosaics from houses
in Conimbriga, Portugal (ibid., cat. nos. 46, 47). A vanquished and dy-
ing Minotaur appears solo in a mosaic from a house in Calvatone, Italy
(ibid., cat. no. 25), and also in a tomb mosaic from Sousse, Tunisia
(ibid., cat. no. 54). I have not been able to find a plan of the small
baths of Hippo Regius, which would allow a reconstruction of the be-
holder’s access to the labyrinth; did he or she enter the room from the
same side that the ball of thread enters the Minotaur’s inner sanctum?
70. Alois Riegl, “Spätrömisch oder Orientalisch?” Beilage zur Münchener
Allgemeinen Zeitung 93–94 (April 23, 1902), translated as “Late Roman
or Oriental?” in German Essays on Art History: Winckelmann, Burckhardt,
Panofsky, and Others, ed. Gert Schiff, German Library Series, vol. 79
(New York: Continuum, 1988), 181. This essay summarizes Riegl’s argu-
ments from his book Late Roman Art Industry, which first appeared in
1901.
71. See Kern, Through the Labyrinth, cat. no. 160. Also Daszewski, La
mosaı̈que de Thésée, 116 –17. Both authors date the mosaic to ca. 50 CE,
without going into detail. Its original dimensions are unknown. Discov-
ered in the eighteenth or early nineteenth century and subsequently
entirely lost, it may have come from a house.
72. Gardiner, “Wrestling,” 16, discussion of Lucian, Hermotimus 40.
73. Other labyrinth mosaics also directly hint at athletic victory: in the laby-
rinth from the Baths at Verdes, France, the central motif is a crown of
laurel leaves. See Daszewski, La mosaı̈que de Thésée, 108, cat. no. 18. Slim
(“La mosaı̈que du labyrinthe,” 209 n. 1) suggested that in this case
there is no relation between the center motif and the geometric frame
of the labyrinth. But the link between the labyrinth and athletic events
can once again render the conjunction meaningful; laurel wreaths
were awarded as victory crowns, such as the one pictured in Fig. 8.
74. The work of Zanker and Ewald (Mit Mythen Leben) on the ways Roman
 ROMAN LABYRINTH MOSAICS AND THE EXPERIENCE OF MOTION
sarcophagi encouraged identification of the self (and the deceased)
with mythological figures is again a productive parallel here. See also
Raeck, Modernisierte Mythen; and Newby, “Greek Athletics.”
75. For the classification, see Yegül, Baths and Bathing, 186 –217. For the
architecture, Thébert, Thermes romaines, 144 – 45. Also Gilbert-Charles
Picard, “Les grands thermes à Mactar,” Bulletin Archéologique du Comité
des Travaux Historiques et Scientifiques, 1972: 151–99; idem, “Les fouilles
de la mission franco-tunisienne à Mactar en 1970 –71: Les grands
thermes orientaux,” Bulletin Archéologique du Comité des Travaux Histo-
riques et Scientifiques, fasc. 8-B (1972): 149 –53. Gilbert-Charles Picard et
al., Recherches archéologiques franco-tunisiennes à Mactar, vol. 1, La Maison
de Vénus (Rome: L’École Française de Rome, 1977), 29 n. 61, labels all
the floor mosaics here as “nonfigural,” though, as we will see, this is
not entirely the case. The authors also note the presence of numerous
tesserae found in the excavations that reveal the vaults were once cov-
ered in mosaic, though of what subject we have no idea.
76. Fikret Yegül (personal communication) feels that these two palaestrae
at Mactar could well have been used for recreational athletics. The two
open courtyards at the Great East Baths of Mactar are repeatedly de-
scribed as palaestrae. See Thébert, Thermes romaines, 144; Broise and
Thébert, Recherches; and Yegül, Baths and Bathing, 196 –97.
77. Plutarch, Theseus 19.2, trans. Perrin.
78. Lucian, Anacharsis, or Athletics 2, secs. 1–2 and 28, have much discussion
of mud and sand and men making themselves either muddy and slip-
Downloaded by [University of West Florida] at 11:37 28 December 2014
pery or sandy and gritty to evade the holds of their opponents. On lo-
cations for wrestling, see Ranon Katzoff, “Where Did the Greeks of the
Roman Period Practice Wrestling?” American Journal of Archaeology 90
(1986): 437– 40. “Why do you roll in the sand?” asks Tertullian in his
attack on the vanity of the palaestra (De pallio 4). In a walkover in a
wrestling match, where an opponent did not show up or simply with-
drew, the victor was said to have won akonitei (without the dust). See
Poliakoff, Combat Sports, 12 and 166 n. 14 with many sources on the
skamma. See also Gardiner, “Wrestling,” 73–74, and “Wrestling (Contin-
ued),” 16 –18. The sandy surface of the wrestling ground also recalls
the single strip of setting given beneath the feet of the combatants pic-
tured at the center of the Thuburbo labyrinth; this, too, was the color
of sand.
79. Alois Riegl, “Excerpts from ‘The Dutch Group Portrait,’ ” trans. Benja-
min Binstock, October, no. 74 (Autumn 1995): 3. This subjective “exter-
nal unity” or “external coherence [äussere Einheit]” was explored by
Riegl in a discussion prompted by intermediary figures glancing out
from paintings to meet the eyes of the beholder; Riegl, “Das Holländi-
sche Gruppenporträt,” Jahrbuch der Kunsthistorischen Sammlungen des
Allerhöchsten Kaiserhauses 23 (1902): 71–278. This has been most re-
cently translated by Evelyn M. Kain and David Britt, The Group Portrai-
ture of Holland, Alois Riegl (Los Angeles: Getty Research Institute for the
History of Art and the Humanities, 1999). See also Margaret Olin,
303
“Forms of Respect: Alois Riegl’s Concept of Attentiveness,” Art Bulletin
71 (1989): 285–99.
80. A parallel to this fusion between decor and experience has been ex-
plored in the realm of Pompeian architecture and painting by Verity
Platt, who explores images of Narcissus and Diana, among others,
whose stories rely on water featuring as part of the beholder’s immedi-
ate ambient; Platt, “Viewing, Desiring, Believing: Confronting the Di-
vine in a Pompeian House,” Art History 25, no. 1 (February 2002): 87–
112. See also Newby, “Greek Athletics,” on unbordered mosaics of
human athletes in Roman baths, especially those deliberately placed in
areas close to the palaestra to blur “the line between bathers and the
mosaic figures” (191).
81. Gilbert-Charles Picard, “Un bas-relief agonistique à Mactar,” Bulletin
Archéologique du Comité des Travaux Historiques et Scientifiques, fasc. 18-B
(1988): 99. But see n. 76 above, for Yegül’s willingness to interpret
Mactar’s two palaestrae as being appropriately scaled to recreational,
rather than professional, athletics. Semipublic athletic competitions at
baths (in Italy) are discussed by Newby, “Greek Athletics.”
82. Picard, “Un bas-relief agonistique,” 98. See also Thébert, Thermes ro-
maines, 455, who says the scene on the bas-relief must have been con-
tinued on another block of stone.
83. For the connection between Eros and athletics, see Thomas F. Scanlon,
Eros and Greek Athletics (New York: Oxford University Press, 2002).
84. Catullus 64.112–15, trans. F. W. Cornish (Cambridge, Mass.: Harvard
University Press, 1976).
85. Tertullian, De spectaculis 18, trans. T. R. Glover (Cambridge, Mass.: Har-
vard University Press, 1931).
86. Natural History 36.19.85.
87. As, for example, the Great Baths at Lambèse, Algeria. See J. Bayet, “Les
statues d’Hercule des Grands Thermes de Lambèse,” in Idéologie et plas-
tique (Rome: École Française de Rome, 1974), 405–10. Miranda Marvin
notes that Hercules is “ubiquitous in bath sculpture”; Marvin, “Free-
standing Sculptures from the Baths of Caracalla,” American Journal of
Archaeology 87 (1983): 379. For the athlete mosaics from the Baths of
Caracalla at Rome, see n. 67 above. Yegül (Baths and Bathing, 175–77)
discusses the College of Herculean Athletes in the Thermae of Trajan
and reproduces an image from the Baths of Nero at Rome, featuring
athletic scenes on a capital today in the Belvedere Court of the Vati-
can. Hubertus Manderscheid offers images and discussions of statues of
athletes from Roman baths in Trier (cat. no. 10), two baths in Ostia
(cat. nos. 82, 98), two baths in Ephesus (cat. nos. 163, 190), and Mi-
letus (cat. no. 225); Manderscheid, Die Skulpturenausstattung der Kaiser-
zeitlichen Thermenanlagen (Berlin: Gebrüder Mann Verlag, 1981). For
mosaics of athletes in baths of North Africa, see esp. n. 41 above.