"Private" Liturgy in Byzantine Architecture :

Toward a Re-appraisal
by Thomas F. MATHEWS

The liturgy is defined as the official public wor- Alongside these changes in planning there
ship of the Christian community, conducted by the occurred a sometimes dramatic change in scale;
appointed ministers of the church on behalf of all for example, the worshipper entering the nave of
the faithful. The symbolism of this ritual the grand Lechaion basilica outside of Corinth
remained fairly constant, apart from certain would have found himself over 100 meters from
nuances, from beginning to end of Byzantine histo- the bishop's throne in the apse; in none of the
r y ; the commentators all understood the identity medieval churches of Kastoria can one find a
of the Eucharist with Christ's own offering of span as great as ten meters from entrance to
himself in his death on the cross and all were apse2. This phenomenon of "miniaturization" of
conscious of the universal efficacy of this offer- church design was by no means universal in
ing'. Yet while the symbolism of the ritual medieval Byzantium, yet it is common enough to
remained constant its external and visible shape require some explanation. How the change in
underwent considerable evolution. The question scale relates to the change in plan is a question
is to what extent can one explain the development that has never been examined. These changes
of medieval Byzantine architecture as a molding of deserve to be examined together for both can be
the architectural shell to fit the ceremonial it tied to similar developments in the evolution of the
housed. Byzantine liturgy. The evidence is far from
The evolution of Byzantine church building complete, and it is perhaps unwarranted to speak
from early to medieval times can be described in of the liturgical changes as "causing" the architec-
the most general terms as an evolution from open tural changes. Still, architecture and liturgy in
to closed forms. The early basilica consisted of a some aspects present parallels so close that one
succession of expansive spaces opening into one must be said to somehow "explain" the other; in
another and lighted from all sides : atrium, aisles, other aspects the connections are at least sugges-
galleries, and nave. By contrast, the medieval tive of areas needing further research.
church was compact and introverted. Whether
cruciform, square or octagonal, its focus was the
well of light in the middle of the building created
by the central dome. Accessibility of space
tended to be noticeably restricted. The fifth cen- The early Byzantine liturgy was very markedly
tury Stoudios basilica, for example, was accessible a liturgy of processions. It is clear that the First
from all sides, even through entrances at the east Entrance was originally a ceremony in which all
to the right and left of the sanctuary ;the medieval the faithful took part; it was the solemn entrance
church, on the other hand, tended to be far more of the public into the church under the leadership
restrictive, often with entrances only on the of the bishop and other clergy j. It began with the
west. Most significant of all were the changes in reception and acclamation of the bishop in the
the layout of the sanctuary. While in the early atrium4, and it ended with the placement of the
Byzantine church the sanctuary extended forward Gospel on the altar and the enthronement of the
from the apse into the nave in a n-shaped plan, bishop in his place in the apse. Similarly, the
with an ambo projecting further into the center of second, or Great Entrance, which marked the start
the church, the medieval sanctuary retreated from of the liturgy of the faithful, was also a true entry,
the nave. The sacred area was organized instead for it was a carrying in of the bread and wine from
in three bays in the eastern end of the church, and a sacristy (or skeuophylakion) located somewhere
the templon barrier that separated the zone of the outside of the church5. The deacons carried in
clergy from the world of the laity ran in a straight the holy elements with incense and candles, and
north-south line. the faithful prostrated themselves, until they
delivered their burden to the celebrating priest at contrary to earlier beliefs, the sanctuary of the
the altar. Further processional action was Early Byzantine church, whether enclosed by a
involved in the readings, in the reception of Com- low chancel barrier or by a colonnaded templon,
munion, and in the recessional or exit of the bish- was perfectly visible to the laity in the
op. churchY. With the passage of time the barrier
By the tenth century all of these processions had between clergy and laity became increasingly
been sharply curtailed. The liturgy no longer opaque eventually evolving into the solid icon-
opened with the procession of the First Entrance screen, a process which took place more slowly
but with a litany that had been prefixed to the than is usually imagined''. Curtains too were
entire ceremony. At the end of the litany the introduced relatively late. The earliest reference,
clergy made an appearance, emerging from the in the Protheoria of Nicholas of Andida (1054-
sanctuary to show themselves and the Gospel, and 1067), describes the shutting of the sanctuary door
returning back to their place in the sanc- with a curtain as a monastic practice; a second
tuary '. By the fourteenth century the commenta- reference later in the eleventh century by Nichetas
tor Nicholas Cabasilas no longer even referred to the Chartophylax describes the enclosing of the
this ceremony as an "entrance" but called it sim- sanctuary with a curtain as a non-Constan-
ply a "showing of the Gospel" (hva6~tSy.rob 'A- tinopolitan custom ".
yiov Edayy~hiou)'. A similar abbreviation took But if the practice was just gaining ground in
place in the evolution of the Great En- the late eleventh century, the mentality that pro-
trance. The preparation of the bread and wine no duced the closed sanctuary was already being pro-
longer took place outside the church but in a moted. As R. Taft has pointed out, Nicetas Stetha-
prothesis chamber immediately to the left of the tos of Stoudios proposed that the laity should
bema. Accordingly' the Great Entrance, though avert their gaze during the anaphora and not cast
still performed with considerable solemnity, was their "unsanctified glance" on the mystery which
reduced to a transference of the elements from the the clergy are performing in the sanctuary 12. In
north bay of the triple sanctuary to the center this instance the change in attitude toward the
bay8. The exact course followed by the ministers liturgy clearly preceded the change in architec-
in transferring the elements from the prothesis ture; the notion that the mystery was too holy for
chamber to the altar is not documented in literary the laity to behold prompted the closing of the
sources, but given the design of Middle Byzantine chancel icreen. What had earlier been a public
churches it probably followed a "U" course out to action taking place visibly in the midst of the
the center of the nave and back to the sanctuary. congregation became now strictly the business of
The transformati~nof the processional entries the clergy shut off in the eastern end of the church,
into "showings" neatly parallels the abandonment and as far as its visible shape was concerned the
of the processional lines of early Byzantine emphasis now.fell on a series of apparitions as the
architecture in favor of the compact, introspective clergy emerged from and returned to their closed
plans of medieval Byzantine architecture. The precinct in the eastern end of the church.
focus of attention in the architecture shifted dra- Closely associated with this re-design of the
matically. The linear design of the Early Byzan- sanctuary is the disappearance of the elevated syn-
tine basilica focused one's attention with compel- thronon. In the Early Byzantine church the cler-
ling force on the apse, the place of the bishop and gy occupied seats in the apse on top a stepped
the altar in front of him ; the medieval church with amphitheatre with the throne of the bishop in the
its central plan and central lighting focused atten- center. Archeological evidence in Constantino-
tion on the door to the sanctuary and the area ple witnesses to this arrangement at the Stoudios +
immediately in front of it, the place of the medie- basilica, Hag. Eirine and Hag. Euphemia, and liter-
val liturgical appearances. The re-organization ary sources describe a seven-step amphitheatre top-
of the sanctuary was designed to suit the restricted ped with a synthronon of silver at Hagia
movement of the medieval liturgy, and the Sophia13. In Middle Byzantine churches the
architect, freed from the necessity of planning his synthronon is brought down to the level of the
structure around a series of parades, turned his sanctuary. In the Pantocrator, an exceptionally
attention to problems in the formal-geometry of large foundation of the twelfth century, it was
symmetrical designs. raised one step above the level of the sanctuary,
Closely linked to this re-focusing of the church but elsewhere it was reduced even further 14. The
plan was the development of an opaque chancel only synthronon in the rock-cut churches of Cap-
barrier. Present evidence now indicates that, padocia, at the tenth century Tokali Kilise, is mer-
ely a pseudo-synthronon consisting of a kind of nion was common only two or three times a
shelf aboui 15 cm. deep (fig. 12). Clearly this year. The liturgy was gradually made ever more
was not intended for sitting; instead, seating for remote, untouchable, inaccessible, invisible. At
two presbyters was provided in carved arm-chairs the same time something else is happening that in
at the right and left corners of the apse a way may be a counter-balance to this developing
(fig. 10). The reduction of the synthronon to two chasm between laity and liturgy, and that is the
seats at Tokali obviously involves a decline in the new intimacy in the dimensions of the church
number of presbyters concelebrating the li- building itself.
turgy 15. On the .other hand the general reduced
importance of the bishop's throne in Middle
Byzantine architecture has its liturgical parallel in REDUCTION IN SCALEAND TENDENCIES
TOWARD
the reduced importance of preaching. "PRIVATE"LITURGY.
In Early Byzantine times the throne, besides
being the symbol of the bishop's position of autho- The general diminution, and at times "miniatur-
rity, was the usual place from which he taught the ization" of medieval church design takes us into
faithful lb. As Krumbacher pointed out, sermons an area of interaction between liturgy and architec-
played a much smaller role in medieval Byzantine ture that remains for the most part unexplo-
literature than they did in early times, and their red. The reduced scale was clearly not a feature
character generally changed from the earlier of the gross decline in population; a town like
homiletic mode of day-by-day commentary on the medieval Kastoria, with its dozens of tiny
scriptural readings to eulogies and encomia for churches, certainly had a total population large,
special occasions ". The decline of spontaneous enough to justify one or two churches of decent
preaching must be linked to the introduction of the size, but this solution was not cho-
"read" sermon. The Quinisext Council (692) sen. Krautheimer and Mango have both attri-
provided impetus for this custom, for while it buted the phenomenon of smaller medieval
reminded the presiding cleric of his responsibility churches to monasticism, and this is probably
in instructing the faithful, it urged him that this basically correct 22 ; inost innovation in church
duty could best be fulfilled by reading from the practice in Byzantium came out of monasti-
Fathers of the Church1'. cism. The question is the mechanism by which
At the same time, the ceremonies of reading this change took' place.
from Scripture and the physical furnishings for the One aspect of this problem, the multiplication
readings underwent a parallel evolution. The of side chapels in medieval Byzantine architecture,
Early Byzantine arrangement called for an ambo has received some attention from
erected prominently toward the center of the nave BabiEL3. Unfortunately BabiC limited her atten-
and connected to the sanctuary by a solea or tion to the mortuary functions of such chapels,
reserved passageway IY. The successive readings whether as martyria, as chapels for the burial of
of Old Testament, Epistle, and Gospel were distinguished founders, or as chapels with other
accomplished with successive processions of the funerary associations. While such uses account
readers along the solea, and the clamor with which for the dedication of a certain number of side
the people surged around the solea to touch the chapels they tell us little about the actual functio-
Gospel after the deacon had finished reading ning of the chapels, which are often equipped in
reminded Paul the Silentiary of the waves of the standard fashion for the celebration of the Eucha-
ocean pounding on a peninsula2'. Again the rist. Moreover a great many side chapels have
medieval Byzantine liturgy transformed the no discernible mortuary associations.
processions into appearances. The ambo disap- The church of the Mother-of-God of Constan-
peared entirely and instead the readings were done tine Lips in Constantinople (907) is an important
on the step, which was now named the solea, case in point. At the east end of the church to the
before the royal door of the sanctuary. The book north and south of the standard triple sanctuary
was placed on a portable wooden lectern2'. arrangement archeology has confirmed the
All of these phenomena point to a general with- existence of two e x t r a side c h a p e l s
drawal of liturgical action from the nave and its (fig. 1) 14. Unfortunately their original furnish-
consolidation within the sanctuary. The liturgy ings were not preserved. On the gallery level,
became more and more a performance of the cler- however, where burials would have been impos-
gy ; lay participation tended to be restricted to res- sible, four more diminutive chapels survive, three
ponse in song, since even the reception of Commu- of which preserved portions of the chancel stylo-

THOMAS F. MATHEWS - aPRIVATEr LITURGY IN BYZANTINE ARCHITECTURE 127

I . Church of the
Mother-of-God of
Lips, Constantinople.
Side chapel to the
south and gallery
chapels. Photo
Mathews.
2. Church of the
Mother-of-God of
Lips, Constantinople.
Gallery plan by E.
Mamboury. Courtesy
Dumbartoil Oaks.
3 . Church of the
Mother-of-God of
Lips, Constantinople.
S o u t h e a s t gallery
chapel. Courtesy
Dumbarton Oaks
Field Committee.

THOMAS F. MATHEWS - rPRIVATEn LITURGY IN BYZANTINE ARCHITECTURE 129

4 . Church of St. Catherine's Monastery, Mt. Sinai. 5 . Church of St. Catherine's Monastery, Mt. Sinai.
Plan. Courtesy G.H. Forsyth. Chapel K . Courtesy G.H. Forsyth.

bates, and two the altar emplacements against the (fig. 4)'?. Structurally all of these belong to
apse with cruciform reliquary settings in the wall the original fabric of the Justinianic
(fig. 2-3)". Two of these chapels were located in church2'. Chambers G and H, furnished with
the western corqer bays where they could be cabinet niches, Forsyth has identified as sacristies ;
reached from the gallery over the narthex, but the the others, however, all have the essential and
other two, in the eastern end of the church, could minimal furnishings for the celebration of the
be reached only through some sort of catwalk divine liturgy. In chapels J and K the altar is
along the side of the church. Their limited access free-standing in a small apse while in the other
and their extremely limited dimensions seem to chapels the altar simply consists of a shelf built
make them hardly suitable for the conduct of the into a niche (fig. 5-7)28. In each chapel another
liturgy, which by its nature is meant to be a "pu- niche immediately to the left of the altar is pro-
blic" action. Yet they contained in miniature the vided for the prothesis ceremony. The finish .of
two most essential elements of the regular celebra- the masonry guarantees the sixth-century date of
tion of the liturgy, namely a barrier setting aside these altar and prothesis niches, for since the
the sacred area and an altar. The archeology building's masonry is dressed only on the surface +

raises the question whether a kind of "private" one could not make a subsequent cut into it
liturgy - a common phenomenon in the West - without exposing the rubble corezY. In each
may have had a role in the Byzantine rite as well. chapel a step indicates the probable location of a
The phenomenon observed at the Lips church, chancel barrier (in N and Q the step is to the side,
namely the multiplication of diminutive chapels from the aisle). The largest, namely chapels J
with the minimal furnishings necessary for the and K flanking the main apse, are about 5 m
celebration of the liturgy of the Eucharist, can be square; those that flank the aisles are a bare 2 m
traced at least as far back as the sixth cen- wide, though of varying lengths. In other words,
tury. A t the monastery of St. Catherine's, Mt. Si- although they are furnished for the celebration of
nai, 548-560, the church follows a common Early the divine liturgy none is large enough to accom-
Christian basilica plan with the notable addition modate a celebration that one could call "public"
of a series of chapels around the periphery or even a celebration that would involve any
6 . Church of St. Catherine's Monastery, Mt. Sinai. 7 . Church of St. Catherine5 Monastery, Mt. Sinai.
Chapel P. Courtesy G.H. Forsyth. Chapel N . Courtesy G.H. Forsyth.

significant percentage of the monastic commu- ready cited the .seating arrangement in the main
nity. One other chapel at Sinai deserves to be apse at the Tokali Kilise. The arrangement in the
mentioned in this connection, an even more dimi- side apses is even more significant for our investi-
nutive chapel located in the center bay of the gation. Both the north and south apse are identi-
southwest wall (fig. 8-9) jO. Slightly less than a cal in furnishings, containing an altar in the center
meter in width and 3.7 m in length, the chapel was and a seat for the presbyter to the right
handsomely frescoed throughout. The altar is (fig. 10). But it should be noted that while the
missing, but the un-frescoed area beneath the cross prothesis ceremony requires a table of some sort it
in the apse probably indicates its location directly has no use for a seat, for there are no scriptural
against the apse wall. Two niches are found in readings; on the other hand, in a diaconicon,
the left wall, the second one placed at a height sui- defined as a place for vessels and for vesting,
table for use for prothesis. The erection of neither altar nor seat would have any place. His-
modern accommodations for the monks along the torians have tended automatically to label the
inside of this wall makes it purely speculative what three sanctuary chambers, wherever they occur, as
relationship this chapel might have borne to the prothesis, berna and diaconicon; the archeology
rest of the monastery plan ; its separation from the suggests, however, that what we have at Tok-ali is
main church and its miniature dimensions are not a single berna with its auxiliary chambers but
significant, however. three bernas side by side, each fully equipped for
We have some evidence, therefore, both at Sinai the celebration of mass. For each bema aprothe-
and at the Lips church pointing toward a "privati- sis niche was provided just to the left of each apse
zation" of the Byzantine liturgy. The develop- in the corridor of space immediately in front of the
ment seems to have been fairly widespread in apses (fig. 10 and 1I). A slightly different
medieval times. Much valuable evidence for the arrangement maintains in the chapel just to the
shape of the medieval Byzantine liturgy is still north of the main church of Tokali. Here the
contained in the rock-cut churches of Cappadocia, chapel has but a single apse. The altar and the
though in their search for frescoes scholars have presbyter's seat to the right are arranged as in the
overlooked the liturgical evidence. We have al- apses of the main church, but the prothesis niche is

THOMAS F. MATHEWS - aPRIVATEm LITURGY IN BYZANTINE ARCHITECTURE 131

 n o r t h chapel "8-
lk,,

8 . Southwest wall chapel,

St. Catherine's Monas-
tery, Mt. Sinai. Plan.
Courtesy G.H. Forsyth. ..
..
...
,
.: . . ..
.. .. ..
:. i. o l d church ji. i.
9 . Southwest wall chapel,
St. Catherine's Monas-
tery, Mt. Sinai. View
toward the apse.
Courtesy G . H . Forsyth.
10. Tokali Kilise, Goreme.
Plan. From G . de Jer-
phanion, Les eglises
r u p e s t r e s , I, 2 ,
plate 61. 8-bemo P - p ~ o t h e r ~ an i c h e S-presbyteri 5eat 1-bishop's throne
Tokali Kilise, Goreme.
Interior, showing alter-
nation of apses (bemas)
and prothesis niches.
Photo Mathews.
Tokali Kilise, Goreme.
Center apse, showing
bishop's throne, pseudo-
synthronon, and pres-
b y t e r ' s chair. Photo
Mathews.

THOMAS F. MATHEWS - nPRIVATEn LITURGY IN BYZANTINE ARCHITECTURE 133

13. Church of Christ of Chdra, Constantinople. Apse and Jlankini chapels. Photo Mathews.

outside of the chancel barrier to the left, in the first liotissa) 31. All of these are diminutive churches
bay of the nave. with a nave and sanctuary spanning less than
Both of these arrangements, with minor varia- 10 m. The two former must have accommodated
tions, are standard throughout C a p p a d o - the prothesis in the niche just left of 'the apse
cia. Where the church has but a single apse the within the chancel barrier ; the third church, which
protehesis is usually in the nave immediately to the was a quatrefoil plan with three niches in each
left of the sanctuary; where the church has three apse, may have accommodated the prothesis either
apses, each apse is a bema with its own altar and in the left niche of the main apse, or somewhere in
presbyter's seat. The commonest variation is in the north conch. The three-bema arrangement of
the placement of the prothesis niche ;if the church Cappadocia also appears in Constantinople at the
is less spaciously laid out than the Tokali a small famous church of Christ of Chora (fig. 13-
prothesis niche is generally placed within the bema 14). Here the main apse has a niche off-center to
itself, to the left, opposite the presbyter's seat. the left, evidently for the prothesis 32, while each of
These Cappadocian phenomena cannot be dis- . the flanking chapels is equipped with apse and
missed as provincial monastic practices; for niche. The fact that the south flanking chapel
although many of the furnishings are missing, the does not communicate with the main apse bears
same arrangements seem to have been in use in out its independence as a separate bema; on the
Constantinople. Churches with a single-apsed other hand the fact that the north flanking chapel
sanctuary are not uncommon; one can cite, for does not communicate with the nave underlines its
example, the Toklu dede Mescidi, the Bogdan unsuitability for prothesis functions in relation to
Sarayi, and the Theotokos Panagiotissa (Moug- the main apse. Beyond the Lips church, there-
 porary, Leontius of Neapolis, narrates the manner
in which St. John sought to correct a government
official who could not be persuaded to be recon-
ciled with a personal enemy.
"One day the Saint sent and had him fetched on
the pretext of some public business, and as soon as
he had come the Patriarch held a service in his
oratory, no one else being present save his syncel-
lus. qfter the Patriarch had said the prayer of
consecration and had pronounced the opening
words of the Lord's Prayer, the three of them
began to repeat the Prayer. When they got to the
sentence : 'Forgive us our debts as we forgive our
debtors,' the Patriarch made a sign to the syncellus
to stop, and he himself stopped, too, and the
magistrate commenced saying all by himself 'For-
give us as we forgive33."'
The situation could not be clearer. We are
dealing with a personal oratory (the text says bv
r@~ I J ~ q p i caljroc),
p probably located within the
episcopal residence (for reasons we will see in the
next passage), and the bishop is celebrating the
divine liturgy with only his secretary (in the role
ofdeacon, one presumes) and the magis-
trate. Beyond this we can also assume, and this
is significant, that we are dealing with a weekday
celebration, since on Sundays and holidays the
bishop would have been expected to celebrate
publicly for his parishioners.
14. Church of Christ of Chdra, Constantinopie. A second refkrence to the private celebration of
South finking chapel, showing apse and pro- the liturgy is found in a passage where the Saint
thesis niche. Photo Mathews. deals with certain parishioners who were in the
habit of leaving the public, cathedral liturgy after
the Gospel to stand outside and talk.
fore, there is considerable evidence in Constanti- "Directly after the Gospel had been read in the
nople that the liturgy was being celebrated on a church, (the Saint) slipped away and came out
reduced, private scale, whether in churches of himself and s a t down outside with the
miniature size or in small chapels annexed to crowd. And when everybody was amazed, the
larger churches. Further monuments to be just man said to them :'Children, where the sheep
considered in the same category might be the are, there also the shepherd must be. Come
gallery-level chapels of Hag. Eireni and the Giil inside and I will come in ; or stay here and I will
Camii, and the multiplied east-end chapels on the stay, too. For I come down to the Holy Church
ground level at Theotokos Pammakaristos, the for your sakes, since I could hold the service for
Kalenderhane Camii, and the Vefa Kilise Camii. myself in the bishop's house34."'
Literary evidence for the privatization of the Clearly the bishop had an oratory within his
Byzantine liturgy is still scarce, but it is hoped own residence (6v r@imo~onsicp)for his private
that by drawing attention to some of the docu- use. Moreover, while the bishop was most
ments further sources can be brought to anxious that as large a public as possible should
light. Curiously enough, the earliest literary attend the cathedral liturgy, he takes the extraordi-
evidence comes from Sinai's patriarchal see of nary position that the congregation is entirely
Alexandria, only a generation or two after Justi- dispensible as far as the essential function of the
nian's church. The evidence is found in the life of liturgy is concerned. Evidently when the duties
St. John the Almsgiver, appointed by Heraclius in of caring for his flock did not require a public cele-
610 as the orthodox, i.e. Chalcedonian, bishop of bration he felt free to celebrate privately.
Alexandria. The saint's biographer and contem- One might assume from these two texts that we

THOMAS F. MATHEWS - rPRIVATEr LITURGY IN BYZANTINE ARCHITECTURE 135

are dealing with an episcopal privilege, but a third purple linens interwoven with gold and treasures
text speaks of monastic priests celebrating in the in abundance3'." Even allowing for the author's
privacy of their cells. tendency to exaggerate, the casual reference is
He collected two bodies of holy monks, valuable information, for it not only documents
arranged "that all their needs should be supplied the existence of oratories in the galleries of Hagia
from the lands belonging to him in his native city, Sophia, but the mention of the gifts of vessels and
built cells for them and appointed them to the two linens clearly indicates that the oratories were
oratories, the one of our Lady, the holy Mother of equipped for the celebration of the divine li-
God, and the other of St. John, which he had turgy. As in Alexandria we are probably dealing
rebuilt from the foundations. Then he spake to with chapels intended for a weekday liturgy for
the monks beloved of God and said : 'I myself - those whose devotion was not satisfied with a litur-
after God - will take thought for your bodily gy restricted to weekends and holy days. But by
needs, but you must make the salvation of my soul the end of the century the practice received a
your care, so that your evening and night vigils conciliar endorsement that seems to embrace the
may be set to my credit with God. But if you private domestic celebration of the liturgy
celebrate the liturgy in our cells, your own souls regardless of the day. Canon 3 1 of the Quinisex-
will gain the benefit3'.'" tum Council reads : "We decree that those clerics
Here we are dealing with a kind of spiritual who happen to be celebrating the liturgy or baptiz-
profit-sharing plan. Each of the two monasteries ing in domestic oratories in homes should do so
was given a communal oratory in which the with the knowledge of the bishop of the
monks could gather for the chanting of the divine place3'." The canon contains no hint of disap-
office. The plan required that in the chanting of proval of the practice, but only insists that it
the office (the text refers to fi h u p u c ~ )~ a fii V ~ K - belongs under the jurisdiction of the local eccle-
rspivq bypunvia) the monks should pray for the siastical authority. The reference to baptism
bishop. At the same time each monastery had implies that the domestic oratory was being used
individual cells for the monks, and, according to for the full range of liturgical services for the
the bishop's plan, if a monk celebrates the liturgy household to which it was attached. It would
in his cell he may do so for his own personal bene- seem, therefore, that by this time the practice of a
fit (&In v a 62 6v TO{< fipkr~p015Kshhioy hE1roup- private liturgy had gained wide popularity,
yiav no~iloqrs, 6x233 r h v 6perkpwv Eo.ra1 whether c'elebrated in monastic cells, in the bis-
y u ~ h v ) . Strange as it sounds, it is difficult to hop's residence, or in a family chapel. It is in this
avoid the impression that the monks customarily light that one must view the multiplication of little
gathered in their common church for the chanting chapels - sometimes adjoining a larger church
of the divine office, but that for the divine liturgy and sometimes separate - in Sinai, in Cappadocia
they were allowed to go to their separate and in Constantinople. What was probably a
cells. Yet the situation is understandable if one monastic practice in origin paved the way even-
assumes that it is again the weekday liturgy that is tually for the family chapel which radically altered
in question. The Early Christian practice, which the pattern of "churching" the population in
was continued far into the Middle Ages in many medieval Byzantium.
places, restricted the public liturgy to Saturday Byzantine law provided for and promoted the
and Sunday; if we assume that the monks were founding of these private churches ( E ~ K T T ~ ~ 01-
IOI
expected to attend the Saturday and Sunday litur- K O ~ ) ,and this aspect of the sources has been exten-
gy in common we must assume that they were per- 'sively researched by E. Herman ". Not only
mitted to celebrate for their own devotion on other great landowners, but anyone who could afford to
days in private. express his piety in this fashion could erect a
The Life of St. John the Almsgiver informs us church or chapel. Such a church remained the
that in Alexandria a private liturgy was known property of the founder who could sell, lease, or
both in a monastic setting and in the episcopal bequeath it, though he was enjoined from aliena-
residence at the beginning of the seventh cen- ting it from religious purposes once establi-
tury. Evidence from Constantinople later in the shed. The founder could appoint the priest,
century points in a similar direction. In the subject to episcopal approval, but with this right
Patria we are informed that the Patriarch Sergius went the obligation usually of providing for the
(6 10-638) "established all the oratories of Hagia priest's livelihood and the services in the
Sophia in the katichoumena (i.e., galleries), dona- church. Monastic churches too legally fell within
ting to them many gold and silver vessels and the same category of E ~ K T T ~ ~ O?KOL,
IOI and monas-
tic property was also generally in the hands of a deal more about the patterns of family patronage
lay owner, the ~ a p 1 o z t ~ a p t o 5 AS
~ ~ a. result "ca- in relation to church building. In places like
tholic" churches, that is churches immediately Kastoria and Mani it seems clear that the extended
dependent on the bishop, became an ever smaller family was the unit for churching, but the evidence
minority of the total number of churhes; in 15th has never been systematically examined4'. Fi-
century Constantinople only eight churches were nally, in the monastic celebration of the liturgy
counted "catholic" against over two hundred pri- one would like to know more about the day-to-day
vates churches 40. calendar of liturgical observance. In post-
Of course, churches that were private legally did medieval times one reads of the practice of a rota-
not have to be private or miniature in scale. Yet ting liturgy moving daily from chapel to cha-
in fact this was the direction which the evolution pel. In the eighteenth century Allatios tells us,
took. The church became a family or neighbo- "In some monasteries there are as many
rhood possession, and while the liturgy in its per- parecclesia as there are days in the week. In
formance became ever more remote and invisible these, except on Sundays and on the feast-days of
to the laity, in its physical dimension it became saints when it is required of all monks to attend
ever more intimate and domesticated. the services, one of the monks to whom the duty
Present evidence for the privatization of the falls, and who is called the 4P60pa6aptoq, cele-
Byzantine liturgy points to a number of areas of brates the rite - one day in one parecclesium, the
possible research. From the point of view of next day in another. In this way by the time a
archeology much data has gone unobserved sim- week has passed he has celebrated in as many
ply due to an a priori presumption against the parecclesia 42."
existence of a private liturgy in Byzantium. A On the other hand a sixteenth century Russian
more objective review of the material should traveller in Sinai reported, "Le couvent du Sinai
uncover considerable new evidence of the functio- posslde en tout vingt-cinq eglises et chapelles;
ning of Byzantine architecture. Among literary I'office divin est celebre d a n s c h a c u n e
sources the lives of the saints deserve new atten- d'elles 43." To what extent such practices reflect
tion from this point of view, especially since the medieval observances has yet to be investigated.
usual liturgical sources of typica and commenta-
ries are by their nature concerned generally with
the public and official conduct of worship. In
medieval villages one would like to know a great Thomas F. MATHEWS.

NOTES
1. On the literature of the commentaries on the Byzan- 6. On the introduction of litanies and antiphons at the
tine liturgy see Rene Bornert, Les Commentaires byzan- beginning of the liturgy see Juan Mateos, La Celebration
tins de la divine liturgie du Vlle au XVe siecle, Paris, 1966. de la Parole dans la liturgie byzantine, Etude historique,
2. On the Lechaion Basilica see R. Krautheimer, Early Orientalia Christiana Analecta, 19 1, Ronie, 197 1. I11
Christian and Byzantine Architecture, Baltimore, 1965, Constantinople the patriarch continued for some time to
pp. 99-101 ; on the churches of Kastoria see the recent enter the cathedral solemnly from his adjacent living
article by A.W. Epstein, "Middle Byzantine Churches of quarters, letting lesser clergy attend to the opening lita-
Kastoria : Dates and Implications," Art Bulletin, LXII, nies, Mateos, pp. 76-79. Generally, however, the de-
1980, pp. 190-207. velopment of the prothesis ceremony, that is the private
blessing of the bread and wine before the start of the
3. T.F. Mathews, The Early churches of Constanti- liturgy, required the presence of the celebrant in the sanc-
nople :Architecture and Liturgy, University Park, Penn., tuary before the litanies. This order of events is attested
1971, pp. 138-147. to by Gerrnanus in the eighth century. Historia Eccle-
4. In addition to the references in Mathews, Early siastica, ch. 20-22 ; ed. Nilo Borgia, 11 commentario litur-
Churches, pp. 144-1 45, one should note the references gico d i s. Germano, Grottaferrata. 1912, pp. 19-2 1.
to a place for the salutation (npvapqcriq) of the bishop in 7. Nicholas Cabasilas, Commentary, ch. 20, ed. S. Sa-
the atria of St. Sergius and St. Stephen at Gaza. Cori- laville et al., Nicholas Cabasilas, Explication de la Divine
cius, Laudatio Marciani, 1, 2 2 and 11, 33, ed. R. Foerster, Liturgie, Sources Chrktiennes, 4, 2nd ed., Paris, 1967,
Choricii Gonaei Opera, Leipzig, 1929, pp. 8 and 36. p. 146.
5. On all aspects of this ceremony and its evolution 8. Taft, Great Entrance, pp. 1-1 0, 178-203.
see the recent monograph by Robert F.Taft, The Great
Entrance, A History of the Transfer of Gifts and other Pre- 9. Taft, Entrance, pp. 41 3-41 ; Mathew%
a n a ~ h o r a lRites of the Liturqv of St. John Chrvsostom, Churches, pp. 62-1
~ r i e n t a l i aChristiana ~ n a l e c i a ;200, Rome, 1975. 10. W e still lack a thorough history of the de-

THOMAS F. MATHEWS - nPRIVATEr LITURGY IN BYZANTINE ARCHITECTURE ,137

v e l o p m e n t f r o m t e m p l o n screen t o iconosta- 27. G.H. Forsyth. "The Monastery of St. Catherine at
sis. Cf. J. Walter, "The Origins of the Iconostasis," Mt. Sinai : The Church and Fortress of Justinian." Dum-
Eastern Churches Review, 3, 1971, pp. 251-267. The barton Oaks, Papers, XXII, 1968, p. 1 1 and n. 13.
eleventh century churches of Cappadocia sometimes
28. Fig. 5-7 correspond t o plates XCIX-6, XCIII-B,
show an arcade screening the sanctuary. but it is signifi-
and XCII-A in G.H. Forsyth and K. Weitzmann, The
cant that icons do not appear on these arcades even Monastery of St. Catherine at Mount Sinai, The Church
when the rest of the church is covered with frescoes. and Fortress of Justinian, plate volume, Ann Arbor,
1 1. Nicholaus Andidorum, Protheoria, 2 1, PG, 140, Michigan, 1970.
col. 445. For Nichetas see Taft, Great Entrance, p. 4 1 2. 29. 1 am indebted t o Forsyth for this observa-
12. Taft, Great Entrance, pp.411-412. Walter tion. On the masonry technique see his remarks in
assumed a general use of curtains in the Middle Byzan- Dumbarton Oaks Papers, XXII, 1968, p. 7, n. 6.
tine period. Cf., op. cit., p. 258. 30. Fig. 3 0 appears as plate IX-A in Forsyth and
13. On the svnthronon at Has. EirEnE see U. Pesch- Weitzmann, The Monastery of St. Catherine. For other
low, Die lrenenkirch in lstanb2, Untersuchungen zur views of the chapel and its frescoes see plates IX-B,
Architektur. lstanbuler Mitteilunoen. beiheft 18.. TO-. CXXXIV-A and B, CXXV-A t o C, and CXClV to CXCVIII.
bingen, 19'77, pp. 24-25. For the &her churches see 31. T.F. Mathews, The Byzantine Churches of Istan-
Mathews. Early Churches, pp. 26-27, 66, 99. b u l : A Photographic Survey, University Park, Penn.,
14. A Megaw, "Notes on Recent Work of the Byzan- 1976, pp. 36-39, 366-382, with relevant bibliography.
tine Institute in Istanbul," Dumbarton Oaks Papers, XVII, 32. Ibid., plate 8-24.
1963, p. 3 3 6 and fig. A. 33. Leontius of N eapolis, Life of John the Almsgiver,
15. On the evidence for concelebration in the Byzan- ch. 39, ed. H. Gelzer, Leontios' von Neapolis Leben des
tine rite see R. Taft, "Ex Oriente Lux ? Some Reflections Heiligen lohannes des Barmherzigen, Sammlung ausge-
on Eucharistic Concelebration," Worship, LIV, 1980, wahlter kirchen- und dogmengeschichtlicher Quellen-
308-325. I n the nave of Tokali seating is provided for schriften, V, Freiburg i.B. and Leipzig, 1893, pp. 78-79 ;
the chanting of the divine office, which gives us a nice trans. E. Dawes and N.H. Baynes. Three Byzantine Saints,
picture of the size of the monastic community which the Oxford, 1977, p. 247.
church served. The benches along the north, south and 34. Leontius, Life, ch. 4 2 ; ed. Gelzer, pp. 83-84;
west walls seated 1 5 monks at the most (one in each trans. Dawes and Baynes. p. 250.
recess); along the east wall in front of the chancel are
three arm chairs for priests and benches between them 35. Leontius, Life, ch. 4 2 ; ed. Gelzer, p. 8 5 ; trans.
for another t w o or four monks. If necessary another ten Dawes and Baynes. p. 251.
monks could be accommodated in the west nave of the 36. Ps. Codinus, Patria. Konstantinoupoleos, ed.
"old church." One is dealing, then, with a monastic T.Preger, Script. Orig. Const., Leipzig, 1 9 0 6 , 11,
community consisting of three priests and around p. 280. R. Cormack has corrected m y earlier use of this
2 5 monks. Cf. G. de Jerphanion, Une nouvelle province text (Early Churches, p. 129) in a note in his article with
de I'art byzantin : les e&lises rupestres de Cappadoce, E. Hawkins, "The Mosaics of St. Sophia at Istanbul : The
Paris, 1932, 1, 2, pp. 297-376. Rooms above the Southwest Vestibule and Ramp,,"
16. Mathews, Early Churches, pp. 149-1 52. Dumbarton Oaks Papers, XXXI, 1977, 199, n. 37.
17. K. Krumbacher, Geschichte der byzantinischen 37. Quinisext Council, canon 3 1 : 70% i v ~ u ~ r v p i o l q
Literatur, Munich, 1897, 1, pp. 160-1 63. O ~ K O IE~V~ O Vo i ~ i a qTU~XUVOVOI ~ - E I T O U ~ Y O V V T U4~ I j ( l n r i < ~ v -
raq ~ h - q p ~ r o i6x0
y , yviuptlq 70670 n p a r r ~ roc ~ v ~ n r ronov &
18. Quinisext Council, canon 29, Mansi, XI, 952. Cnio~onouopi<op~v.Mansi, XI, 956.
19. For archeological evidence in Constantinople of 38. E. Herman, "Die kirchlichen Einkunfte des byzan-
the solea and ambo see Mathews. Early Churches, tinischen Niederklerus," Orientalia Christiana Periodica,
pp. 36-37, 53-54. 66, 7 0 and 98-99. Vll, 1942, pp. 378-442, esp. 419-425; "'Chiese private'
20. Descriptio ambonis, lines 244-251, ed. P. Fried- e diritto di fondazione negli ultimi secoli dell'impero
lander, Johannes von Gaza und Paulus Silentiarius, Leip- bizantino," Orientalia Christiana Periodica, XII, 1946,
zig, 1 9 12, pp. 263-264. 302-32 1 ; "The Secular Church," in Cambridge Medieval
History, IV, part 2, Cambridge, 1967, pp. 104-1 33,
21. Portable lecterns appear in Byzantine art in
esp. 116-1 25.
representations of Christ reading in the Synagogue.
39. E. Herman, "Richerche sulle istituzioni monas-
22. R. Krautheimer, Early Christian and Byzantine tiche bizantine : typica ktetorika, caristicari e monasteri
Architecture, p. 2 4 7 ; C. Mango, Byzantine Architecture, 'liberi,' " Orientalia Christiana Periodica, 11, 1940, 2 9 3 -
New York, 1974, pp. 197-1 98. 3 7 5 ; P. Lemerle, "Un aspect du r61e des monasteres a
23. G. BabiC, Les Chapelles annexes des eglises Byzance : Les Monasteres donnes a des la'ics, les charisti-
byzantines : Fonction liturgique et programme iconogra- caires," Comptes rendus de I'Acad&mie des inscriptions
phique, Paris, 1969. et belles Lettres, 1967, pp. 9-28.
24. A. Megaw, "The Original Form of the Theotokos 40. E. Herman, Cambridge Medieval History, IV, e
Church of Constantine Lips," Dumbarton Oaks Papers, part 2, p. 1 18.
XVIII, 1964, pp. 279-298. 41. For a general introduction see D.E. Rogan, M a n i :
25. T. Macridy, ''The Monastery of Lips and the History and Monuments, Athens, 1973. In Kastoria
Burials of the Paleologi," Dumbarton Oaks Papers, XVIII, Epstein has remarked on the pattern of family patronage
1964, pp. 253-278 and fig. 24-27. in Art Bulletin, LXII, 1980, 200-1.
26. 1 am deeply grateful t o Prof. George H. Forsyth of 42. Leo Allatios, ch. 7, trans. A. Cutler, The Newer
the University of Michigan for his generosity in sharing Temples of the Greeks, University Park, Penn., 1969,
the results of his archeological study of Sinai before his pp. 34-35.
final publication of the material. In addition t o the pho- 43. S. Khitrowo, Itineraires russes en Orient, Geneva,
tographs of fig. 4-9, which Prof. Forsyth made available 1889, p. 303.
to me, I was also the beneficiary of his patient exposition I am grateful to Father Robert F. Taft, S.J., for calling this
of the material in conversation and correspondence. reference to my attention.