Snapshots on the Sculptural Heritage of the White Monastery at Sohag: The Wall Niches
THE SCULPTURAL REMAINS of both monastic churches in the Sohag region, the so-called Red and White Monasteries (Dayr al-Ahmar and Dayr al-Abyad), is of great importance with regard to our absolutely insufficient knowledge of Coptic architectural sculpture in general.
It is primarily the White Monastery, but also the Red Monastery, that provides a first-rate sculptural heritage in the truest sense of the word. Although both churches contain a mixture of spolia besides the pieces originally made for the building, and both churches suffered much destruction (and these destructions, especially the following repairs, may falsify in some cases the original outlook of the sculptural composition), enough of the original sculpture of the mid-to-late fifth century survived in situ (that is, in its original functional context, if not at its original place), thus allowing to build up an important chronological specification for the sculptural production of the fifth century. Moreover, it is possible to analyze the sculpture within a given architectural context.
In the following, we will concentrate mainly on the White Monastery, because in the Red Monastery by far the most of the ‘sculptural’ decoration were painted as a substitute for carved decoration. Even the relatively few carved architectural members, as, for example, the column or pilaster capitals, the pediments of the wall niches or the supraports, were enriched by a layer of polychrome color—this, however, might have been the rule rather than the exception not only in the classical Greek and Roman world, but also in Egypt from the pharaonic period on to late antiquity. Thanks to the extraordinary well-preserved condition of the interior of the triconch sanctuary in the Red Monastery, we have a vivid impression of this. However, even in the White Monastery, as we will refer to in the following, we can find traces of painting both as a substitute for and as additional adornment of carved decoration.
State of Research on the Sculptural Decoration of the White Monastery
Besides the former more general and almost incidental references to the sculptural decoration in what is until today the major study of the two church buildings, made by Monneret de Villard in 1925-1926, and the questionable, at least, value of the attempt undertaken by Philipp Akermann in 1976, one should primarily refer to the contributions of Hans-Georg Severin. However, his research is in its initial state. We are still lacking the substantial basis for a systematic research of the sculptural program. We have neither a systematic and complete photographical nor an analytical documentation of the church building itself or of its sculptural decoration.
The aim of the present contribution can neither be to propose a complete, distinctive discussion of the sculptural program, nor to repeat sketchy overviews of the different sculptural members. On the contrary, the focus will be on one part of the original architectural conception of the mid-fifth century—the wall niches—whose significance is not only attested by its role within the architectural plan; thanks to the preservation of their rich sculptural decoration, the wall niches become a ‘talking witness’ of the sculptural heritage of the White Monastery.
It is important to note that the following statements present only an intermediate status of work—they are first snapshots of the material.
In order to obtain a concrete idea about the ‘architectural’ significance of the wall niches on the one hand, but also their actual status quo, it is necessary to refer first of all to the building itself, namely a comparison of its original architectural conception and its actual appearance (fig. 1). In its original architectural conception the church was of primarily basilical character, with three aisles and a return aisle on the west side with upper galleries and an elaborated triconch sanctuary at its east end surrounded by several annex rooms, among them an octagonal baptistery. At the western end of the church a narthex with columned exedrae at its lateral ends was built, and along the south side of the church a long, narrow hall with an absidial end on its western side was attached—the so-called ‘south-narthex.’ At its eastern end a door still leads to a square room (the former library), whereas at its western end there was once a rectangular room.
As already mentioned, the church suffered several periods of destruction and repair, the first collapse as early as between the sixth and eighth centuries. After the collapse of the wooden roof of the naos at some time in the Middle Ages, it most probably remained unroofed and lost its function. The rebuilt church was reduced almost to the size of the former sanctuary with the addition of an area in front of it. The remaining part of the former naos was later occupied by several civilian structures that were removed in 1984. In the southwest part of the church the wall between the nave and the south narthex collapsed and the western absidial end of the south narthex disappeared almost completely. Finally, the former rectangular room in the southwest corner of the church was totally rebuilt to a square-domed structure in the beginning of the nineteenth century.
Concerning the original architectural conception, we can clearly state the continuous increase of complexity from west to east. In spite of the later reconstructions and repairs one could state a parallel increase also in the complexity and richness of the sculptural decoration with its climax in the triconch sanctuary with its two-storyed columned architecture separated by broad entablature zones (in the upper story supporting the semidome), and niches arranged in the intercolumnar sections of the wall behind. However, the wall niches as architectural feature are not restricted to the triconch sanctuary. We can already clearly state from the ground plan that they dominate all the main parts of the church except to the square room at the east end of the south narthex.
The wall niches
According to Grossmann’s reconstruction, there are seventy-seven wall niches. The numeration in fig.1 is based on those made by Akermann (Nos. 1-56) and is supplemented by those niches that Akermann did not count (Nos. 60-65, 67-69) respectively that have been reconstructed by Grossmann (Nos. 57-59, 66, 70-77).
The seventy-seven niches are distributed in the building as follows:
- Triconch sanctuary: thirty niches (Nos. 1-30): Each conch contains ten niches, arranged in two superimposed rows.
- Baptistery: six niches (Nos. 35-40)
- Nave: eighteen niches (Nos. 31-34, 41, 42, 50, 51, 58-67):
Niches 59 and 60 have disappeared completely; niche 58 figures only in a small rest of the lower part of the semicircular niche back, and of niche 33 only a small fragment of the niche head is still visible. Niches 51 and 31 are each almost half hidden behind later constructions.
- West narthex: seven niches (Nos. 43-49)
- South narthex: sixteen niches (Nos. 52, 53-56 [each time two niches one upon the other], 57, 68-77); niche 57 is not preserved. Concerning niches 70-77, those within the columned exedra (Nos. 72-76) have disappeared, whereas the status quo of niches 70, 71 and 77 cannot yet be verified.
In the White Monastery we can number at least four different variations of niches (fig. 2[a-d]). To be exact, we can distinguish on the basis of two principal niche types—the barrel-vaulted rectangular niche and the hemispherical-vaulted semicircular niche—between three realizations of the niche crowning:
1. Hemispherical-vaulted semicircular niche crowned by an archivolt (see fig. 2[a])
Owing to the state of preservation of the wall niches and the lack of research in this area, it is difficult to decide whether or not a niche belongs to this first group—this problem occurs also with the following groups (for example, see in this context niches 62, 36, 38 or 39; more certainty perhaps concerning niches 35 and 52). The archivolt of the niche 25, however, may belong to a later ‘repair.’
2. Barrel-vaulted rectangular or hemispherical-vaulted semicircular niches with a broken niche pediment: The broken niche pediment is attached to the niche itself (see fig. 2[b-c])
Strictly speaking, a broken niche pediment is a composite pediment, combining the triangular (for example, niches 8, 10, 28, 30, 32, 47, 48, 53-56) or curved (for example, niche 9) center of the pediment above the niche itself with its two lateral broken ends, the latter resting on supports flanking the niche.
As a rule, one can state for the White Monastery that the choice of the supports depends on the type of niche, that is, the broken niche pediment of a barrel-vaulted rectangular niche rests on half pilasters with half capitals of Corinthian order (see fig. 2[b]), whereas in the case of a hemispherical-vaulted semicircular niche half columns, also with half capitals of Corinthian order, are chosen (see fig. 2[c]). The complete pediment is framed by a cornice. Both niches are further accentuated by a cornice running along the inner surface of the niches forming the lower limitation of the niche head. Beneath this cornice one can notice in few cases a horizontal frieze with a painted decor or at least traces of paint on it—perhaps a substitute for carved decoration.
The broken niche pediment is a distinctive Egyptian architectural feature deriving from Ptolemaic tradition. Concerning the question of its continuation throughout the Roman period until the fourth century A.D., the discussion is still in progress.
3. Barrel-vaulted rectangular or hemispherical-vaulted semicircular niches with a broken niche pediment: The broken niche pediment with a triangular center is separated from the niche as an autonomous member resting on supports, whereas the niche is crowned by an archivolt (see fig. 2[d]).
As far as one can state owing to the problems already mentioned, the niches of this third group constitute the exception in the White Monastery (for example, see niches 45, 46 and 50; problematic are niches 63 and 49). By contrast, in the Red Monastery all the niches in the upper story of the conchs in the triconch sanctuary belong to this third group as further niches at the elaborated entrance front into the sanctuary. Also regarding these niches, the choice of the supports of the niche pediment obviously depends on the type of niche.
The two niche types are arranged, as Grossmann stated, in a strong system of alternating square and semicircular niches both in the horizontal and in the vertical order, also comprising the window and door openings as substitute for square or semicircular niches. Such a system of alternating semicircular and square niches is well known from Roman architecture back to the first century B.C.
D) State of preservation (fig. 3)
Although the niches belong to the original architectural concept of the fifth century and were made of limestone, as were all other parts of the original architectural sculpture (whereas the pieces of granite and marble belong to later repairs or were reused spolia), in different respects they also represent the several periods of destruction and repair. The range of preservation includes the following niches: 1) Remarkably well-preserved niches (for example, niches 9 (see in fig. 2[b]), 10, 28); 2) Niches with essential damage, as in the broken niche pediment and its lateral supports (for example, niches 7, 8, 41, 42, 61); in some cases only the contours of the broken niche pediment remain visible (for example, niches 6, 11, 12, 15, 17-19, 22); or 3) Niches whose original substance seems to be reduced to almost the fragmentary rest of the niche head inserted into a rebuilt (?) niche structure (niches 1-5, 21, 23, 64) or to be found without context in the debris.
The repairs and reconstructions of the niches ‘create’ further problems such as the strange, hybrid appearance of some niches (niches 49, 50 or 63) or the obvious mistakes (see niches 54, 56 and 25 with hemispherical- vaulted niche heads inserted in a false way into rectangular niches).
Because of these and other manipulations of the original material and the lack of analytical research there is no methodical basis for attempts such as that undertaken by Laszlo Torok to create by means of the decoration of the niche heads a theological and/or liturgical concept.
E) Sculptural decoration
1. Niche head
At least fifty-three niches present a niche head with sculptured decoration. The repertoire of motifs comprises the following six topical groups (fig. 4):
- Vegetal patterns (tendrils and branches of various plants): ten niches (1, 3, 5, 7, 9, 16, 27, 29, 37, 39)
- Shells: seventeen niches (4, 6, 11, 15, 25, 30 (with a central cross),
- 35, 38, 47, 48, 50, 52, 53, 55, 65 and the separate fragment of a niche head)
- Vases with growing vine tendrils: eleven niches (10, 14, 17, 22, 24, 28, 31, 36, 40, 51, 56)
- Animals: eight or nine niches (eagle: 32 (with peacocks), 66; peacock: 42, 61, 67; gazelle/antelope: 12, 18, 41; see also 33)
- Wreathed crosses: three or four niches (2, 8, 26; see also 33)
- Interlace (looped) surface patterns: three niches (13, 19, 54)
It appears to be striking that there seem to be no counterparts, despite of the topic they have in common, especially with regard to a relatively homogeneous motif, like the shell, for example.
2. Cornice beneath the niche head
- Modillion: niches 34, 35, 36, 41, 42, 45, 46, 49, 51(?), 52, 65, 66, 67
- Band of acanthus leaves: niche 56
- Zigzag band: niches 9, 10 (?)
- (Undulating) vine tendril: niches 63, 38
- Row of tangent circles formed of four spindels: niche 62
- Two-strand guilloche: niche 37
3. Intrados of the hemispherical-vaulted niche
- Two rows of looped circles with floral filling motifs: niche 66
- Surface pattern: diagonal grid with filling motifs: niche 41
- Two-strand guilloche of vine tendrils: niche 39
- Surface pattern: tangent octagons forming squares with filling motifs: niches 9, 37
- Acanthus scroll enclosing rosettes: niche 61
4. Center of the broken niche pediment (fig. 5)
- Variations of tendrils/branches of different plants (sometimes flanking a central cross): niches 10 (vine tendril), 27, 31, 41, 53, 54(?), 61, 66(?)
- Two-strand tendril forming circular medallions filled with rosettes: niche 67
- Looped guilloche: niche 28
- Two-strand guilloche with an alternation of closed and opened loops, the latter forming circular medallions filled with rosettes: niche 65
- Two-strand guilloche with opened loops, forming circular medallions filled with floral motifs: niche 9
- Central cross flanked by a two strand guilloche with opened loops, forming five circular medallions with different vegetal and ornamental filling motifs: niche 55
Two-strand guilloche with opened loops, forming lozenges with floral, trefoil and quatrefoil filling: niches 2, 56
- Looped zigzag band: separate niche fragment found in the debris
- Row of tangent squares and lozenges with trefoil and quatrefoil filling: niche 47
- Two peacocks drinking from a central vessel: niches 26, 42
- Two gazelles/antelopes (?) flanking a plant (tree?): niche 32
- Row of tangent rectangles with inscribed lozenges and squares in alternation, both with further filling motifs: niche 50
- Two-strand guilloche: niche 35
- Central cross flanked by vases with growing vine tendrils: niche 52
6. Cornice of the broken niche pediment or the archivolt
- Two zones: modillion and astragal/bead and reel: for example, niches 9, 10, 34, 42, 47, 48, 50, 53, 54, 55, 56
- Zigzag band with trefoil filling: niches 2, 30
- Two zones: astragal and two-strand guilloche: niche 32
7. Supports of the broken niche pediment
- Half columns: The preserved examples (for example, niches 8, 10, 26, 28) are decorated with half vertical and diagonal fluting.
- Half pilasters: The preserved examples (niches 9, 27) are decorated with an elaborated motif consisting of a pair of tangent opposed undulating bands/tendrils forming oval medallions filled with floral motifs.
Half capitals: The preserved examples are of Corinthian order.
 Although it is well known that these two buildings are indeed only the churches of the former monasteries of Anba Shinuda and Anba Bishoi, their popular names will be used in the following; see Grossmann 2002b: 528.
 Concerning the situation of Coptic sculpture, see with additional bibliographical references for example, Severin 1977: 243-53; 1981: 315-36; 1991a: 2112-17; 1993: 63-85; 1998: 295-338; Torok 1970: 437-84; Thomas 1989: 54-64; Krumeich 2003: passim.
 For the “Coptic” period, see for example, Thomas 1989: 54-64 (passim); Thomas 2000: passim (cf. Index), esp. pp. 26-27. For some examples of painted sculpture in limestone or wood, see Torok 2005a: 180, 187-88, 207-208, 238.
 See in note 18.
 See Monneret de Villard 1925-1926, vol. 2: 121-31.
 Akermann 1976. He tried to present a catalogue of the preserved and/or visible wall niches with carved decoration and the friezes and cornices of the triconch sanctuary. See the review of Severin 1980: 100-101.
 See cited titles in note 2.
 Our major source for pictures is still Monneret de Villard’s study of the two churches. See in addition to this few published pictures of the courtesy of H.-G.Severin (cf. in Krumeich 2003, vol. 1: figs. 3-5, 26, 31, 32, 43, 44, 47, 50, 52, 54) and the project of a photographical documentation excluding the sanctuary, which was undertaken by E. Bolman (see Bolman 2004a: 381-82).
 See for example, Severin 1991b: 769-70; Grossmann 2002b: 126-28.
 I would especially like to thank Prof. Dr. Siegfried Richter for his immense support in taking the photographs. Except fig.1 (see note 11), the other figures are based on photographs taken by him or the author.
 Fig. 1 attempts to present an approximate idea of the present-day state of preservation of the church with focus on the wall niches. The plan is based on the reconstruction of the original ground plan and a detailed plan of the actual appearance of the sanctuary, both prepared by Grossmann (see Grossmann 2002a: 528-36, fig.150 and Grossmann 1982: 116, fig. 48A) and for the southwestern part of the church on the ground plan published by Akermann 1976.
 See Grossmann 1991a: 768.
 See still some column shafts in situ in a picture published by Monneret de Villard 1925-1926, vol. 1: fig.23. Today the exedra is completely ‘cleaned.’
 The reconstruction of the niches 70 and 72-76 is relatively ‘up to date’ (see Grossmann 2002a: fig.150 dated to 1999) as they are lacking in the previous publications of the reconstructed ground plan of the church (see for example, Grossmann 1984-1985: 71, fig.1; Grossmann 1991a: 768 and Grossmann 1998: 233, fig.13). This important fact is obviously not recognized by Torok (2005b: fig.28) as he refers to the obsolete ground plan of 1998.
 For the different type of niches, see Hornbostel-Huttner 1979: 17.
 Cf. Severin 1993: 63-85 and Krumeich 2003, vol. 1: 125-32.
 The form of the center of the pediment seems not to depend on the type of the niche.
 See especially the niche 65 with a red-colored zigzag band with outline in brown-red color, framed by orange bands; see perhaps also the niches 8 and Compare the possible traces of paint on the horizontal zone beneath the cornice on the wall in the southeast corner of the south-narthex. Further traces of paint are also visible on other architectural members of the niches, so on the cornice beneath the niche head (for example, niches 9, 10, 45, 61, 65); the broken niche pediment (for example, niche 10) and the supports of the broken niche pediment (for example, niche 10).
 See contrary to the negative position of for example, Severin (1993: 70-71) on the base of new finds Torok (2005b: 119-20); cf. also McKenzie 1996: 116.
 See Grossmann 2002b: 118-20.
 See Torok 2005b: 159-62. Torok refers only to the niches, which figure in Akermann’s work, ignoring those niches in the nave which were “uncovered” in the meantime; cf. also note 14.
 The documentation of the sculptural decoration remains incomplete, because several niches cannot yet be studied. Because of the apparent problems of an adequate nomenclature for ornamental motifs (see Hodak 2004: 53), a reference is, whenever possible, made to a comparable motif in the catalogue of Balmelle and Prudhomme 1985.
 According to Akermann (1976: 5 note 3) this niche is without sculptural decoration and was never sculptured as also the niches 15, 20, and 49.
 The interpretation of this motif is difficult; see Akermann 1976: 25.
 See note 24.
 Concerning the etymology of the Greek word for ‘conch’ from the shell decoration itself, see Hornbostel-Huttner 1979: 3-4, 20; for the shell motif, see 195-99.
 According to Akermann 1976: 72-73; actually hidden behind a modern icon placed in the niche.
 See Krumeich 2003, vol. 1: 143-48. According to Torok (2005b: 161-62) it is a triumphal (jeweled) crown held in God’s hand.
 According to Akermann (1976: 20-21) this niche should be decorated with a shell motif.
 According to Akermann (1976: 84-85) the center of the decoration is a wreathed cross.
 See on the contrary Akermann (1976: 22-23), who pretends that the niches 3 and 5 present an almost identical decoration.
 See Krumeich 2003, vol. 1: 115-22.
 Cf. Krumeich 2003, vol. 2: 79-80; Krumeich 2003, vol. 1: pl. 47; the cornice was not noted by Akermann 1976: 130-31.
 See note 18.
 Concerning this motif in general, see Krumeich 2003, vol. 1: 80-96.
 Cf. Balmelle and Prudhomme 1985: 94, pl. 46a-b.
 Cf. ibid.: 120-21, pls.70c-j, 71a-e.
 It was not yet possible to verify the intrados of the niches 1, 3, 5, 12, 14, 21, 23, 27, and 29.
 Cf. ibid.: 368, pl. 235a.d.
 Cf. ibid.: 188-89, pl. 124a-c.
 Cf. ibid.: 251, pl.163b.
 Cf. ibid.: 124-25, pls. 74e.h, 75a.
 Cf. ibid.: 119, pl. 69f-g.
 Cf. ibid.: 119, pl. 69a-b.
 Not noted by Akermann 1976: 68-69.
 According to Torok (2005b: 161) these animals represent ‘harts.’
 Not noted by Akermann 1976: 94-95.
 The modillion zone is not noted by Akermann 1976: 130-31.