From its role in Plato’s Allegory to its invocation in the acronym for Computer Aided Virtual Environments, the cave has been the site of the confrontation of nature, technology, and simulation in various contexts throughout history. Nowhere is this more evident than in the rock-cut settlements found throughout the Mediterranean provinces that were once part of the Byzantine empire. The anachronistic “otherness” of cave architecture invites its interpretation as a skeuomorph-in-reverse: architectural forms stripped of their structural function and used as ornament, albeit in a less technologically advanced medium. However, while the cave can claim to be even more architecturally primitive than Laugier’s archetypal “primitive hut”, the similarity and multiplicity of the Byzantine rock-cut settlements suggest a technological epistemology that bears further investigation. It has been shown that even in the provinces the Byzantines had the technological means to build structures comparable to those of the capital in Constantinople, and that this knowledge was dispersed through a surprisingly peripatetic network of monastic communities. I propose that the repurposing and/or evocation of the cave as architecture in the Byzantine world is insufficiently understood in terms of pragmatism (that it was easier or safer than building) or symbolism (that caves simulated the ascetic desert condition appropriate for monastic life) – explanations typically relied upon by archaeologists and historians. Rather, the proliferation of rock-cut settlements was a deliberate cultural choice that represented a form of provincial resistance and self-determination, which operated though formal modes which I will define in terms of ideas derived from postmodern cultural theory: namely the simulacrum and l’informe (“the formless”).
The use of these terms, inasmuch as they were constructed to comment on the cultural landscape of the twentieth century and not the medieval period, is as much a critique of historical discourse as it is an attempt to formulate new positivist knowledge about Byzantium. Just as archaeologists now use state-of-the-art technology to excavate the empirical palimpsest of a site – still without hope of eradicating uncertainty about the past – it seems equally as valid to deploy whatever critical tools can uncover previously unrevealed layers of interpretation. In order to ground my argument in as much historical context as possible, I will introduce the phenomenon of troglodytism with specific reference to the site of Pantalica in Sicily and the forces at play which would have motivated a concept of territory aimed at resistance to the frequently shifting exchanges of power in the periphery of Byzantium. I will then discuss how cave architecture manifests resistance through the radicality of the simulacrum and the operation of “formlessness”.
Pantalica and Provincial Self-Determination
The site of Pantalica lies on the slopes of a steep promontory in eastern Sicily. The cliffs are honeycombed with prehistoric tombs of the Bronze and Iron Ages, many of which were converted into dwellings and churches in the Byzantine period. The chronology remains a matter of debate, with some interpretations dating the dwellings as contemporaneous with the prehistoric tombs, while others suggest that they could be as late as the post-Byzantine period. In fact both may be true. The erasure of chronology is endemic in the process of reuse, and in the case of Pantalica, may have been the intended result of a particular way of occupying the landscape. The pattern of conversion is puzzlingly inconsistent – the contents and form of some tombs were left intact while others were cleared out and destroyed. While this confounds traditional archaeological notions of cultural continuity, the deliberate choice to reuse rather than excavate new spaces is attested by the distinct topographical relationship between the tombs and the settlements, as is, perhaps, the deliberate choice not to distinguish spatially between sacred and profane, life and death, new and ancient (fig.1).
The inhabitants of the caves are equally elusive. Byzantine troglodytism is often associated with monasticism, and this interpretation was applied wholesale to the entire region of Cappadocia in central Anatolia, until recent studies called it into question. Scholars of the rock-cut settlements in Cappadocia now associate monasticism with identifiable architectural features such as refectories (long tables and benches carved out of the rock). The absence of such refectories in Pantalica suggests that it was not a monastic settlement, though due to the apparent predilection for programmatic ambiguity at the site it is hard to see this as conclusive. What can be confidently surmised, however, is that the area saw the influx of successive waves of cultures from all over the Mediterranean, and that the population was neither autochthonous nor stable. Sicily had been subject to heavy Greek influence and migration since antiquity, and the region had long been accustomed to cultural diversity. By the 10th century, after the Arab incursions, intermarriages between Moslems and Christians were not uncommon, and Byzantine and Latin religious practices were tolerated alongside Islam. It seems likely that the presence of an actual monastic settlement does not preclude the influence of Byzantine monastic ideals on secular culture, given the circulation of information around the Mediterranean, particularly in the guise of individual monks who traveled between communities. The most pervasive element of this monastic ideal, and its coincidence with many troglodytic settlements in other culturally diverse, provincial zones of the Empire, is a certain do-it-yourself communalism that found affirmation in the architectural medium of the cave.
The provincial monastic model asserted the self-reliance of the community. Cave dwellings required less expertise and hierarchy of organization than masonry building, and supported the egalitarian distribution of labor. This does not indicate a rejection of technology: the life of Elias Spelaeotes, a monk from south Italy, recounts that a skilled professional stonemason was recruited to convert a natural cave into a chapel. The stonemason subsequently joined the community and shared the work of outfitting the interior with the brothers and Elias, who was hegumenos (abbot) of the monastery. In practice, this allowed the monks to evade the authority of the local bishop, who had the putative power to grant permission for the construction of new monasteries and the migration of monks between them. This disregard for authority may have appealed to secular communities who chafed under the jurisdiction of foreign rulers or the iconoclastic dogma of the church. Though Pantalica’s physical prominence and inclusion in church documents suggest that it may not have been exempt from taxation or censure, it embodied an act of resistance on another level: the rejection of architecture and the representational modes which constitute architectural culture.
From Skeuomorph to Simulacrum
The erasure of chronology is the means by which the skeuomorphism of rock-cut architectural features attains the radical production of the simulacrum. The skeuomorph performs the first destabilizing procedure – it undermines the causal assumptions embedded in notions of technological progress. It deploys familiarity as a red herring that obscures its subversive agenda while seeming to reinforce comfortable notions of past and future. Like the dislocation of temporality effected through the reuse of caves, skeuomorphs make it difficult to distinguish artifacts from cultural evolution.
The divorce of a formal signifier from its means of production initiates the process towards the simulacrum. Byzantine culture was deeply aware of, if ambivalent about, simulation and its implications. The Iconoclastic Controversy of the 8th and 9th centuries attests to the seriousness of the issue. The political and social effects of iconoclasm were widespread, but iconoduly persisted in the provinces, particularly in the troglodytic regions. A rock-cut church in Cappadocia bears an inscription below an image of Christ that, in its typical English translation, seems to uphold the direct relationship of resemblance between model and copy:
Mikros o typos. Megas o phoron
aperanton typon. Tima ton typon.
This is usually translated as “the image itself is insignificant. But mighty is He who carries upon himself the image of the Infinite. Worship the prototype”. However, there are many problems with this translation. Here typos (and its accusative form typon) is interpreted both as “image” and “prototype”; or, as both model and copy. This maneuver brings the statement in line with a representational model of the image and the divine, and with a similar statement of St.Basil: “the reverence for the icon reverts to its prototype”. An alternative translation, more consistent and less interpolative, reads as:
The image is small. Great is the fear [caused by it]
Seeing the image, venerate the image.
The meaning is less clear, but much more interesting. We have shifted from the realm of representation to the realm of simulation. To say that the image is nothing, and only the fear caused by it is great, is a radical statement. There is no reference to the greatness of the prototype. As Baudrillard suggested, “one could say that the icon worshipers were the most modern minds, the most adventurous, because, in the guise of having God become apparent in the mirror of images, they were already enacting his death and his disappearance in the epiphany of his representations”. According to this logic, if icons didn’t mask the truth that there was no divine model behind the images, why persecute iconoduly so vehemently?
Baudrillard’s interpretation of Iconoclasm became the basis for his definition of the simulacrum. The simulacrum is a copy of a copy, an image that has lost its referent. For Baudrillard, the simulacrum was a symptom of postmodern entropy, the implosion of meaning in an endless network of signs. However, there is no reason to conclude that the simulacrum was cause for despair to the Byzantines of the Cappadocian inscription. The defiance of “the image is meaningless, worship the image” recalls Deleuze’s claim for the simulacrum as a productive, if chaotic force: “the simulacrum is not a degraded copy, rather it contains a positive power which negates both original and copy, both model and reproduction”. Out of this negation comes the creation of a new, hybrid entity that is not subject to existing modes of power and control. Perhaps this is the real reason the Church felt the need to eradicate icon worship – the simulacrum is dangerous to institutionalized cultural production.
It seems reasonable to suppose that many of the monastic and secular communities in the provinces who practiced troglodytism as a form of self-determination were also drawn to iconoduly. The motivation behind the two practices may have been similar. By the ninth century, when icon worship was restored and celebrated as the “Triumph of Orthodoxy”, the Church had found a way to bring the simulacrum under control – to reduce it to the dynamic of model and copy. In architectural production this was reflected in the emergence of the highly conservative cross-in-square diagram that would persist for centuries. The plan and form of the cross-in-square church was replicated throughout Byzantium, each dimension of its interior and exterior coded with liturgical meaning. The individual building, like an icon, was less important in itself than as an image of the prototype. By contrast, cave churches undermined the church-as-image by denying it an outward aspect altogether. The troglodyte carvers tended to favor the single-nave church over the cross-in-square plan; there are no cross-in-square churches in Pantalica, and few in Cappadocia. This appears to be the result of preference rather than ignorance or lack of skill. When the cross-in-square type was used, cupolas were multiplied and structural elements attenuated to allow the skeuomorphic subversion to take hold, to emphasize that this is a copy of a copy, to allow the reassertion of the simulacrum.
The Byzantine concept of the simulacrum was a material one. Typos was understood to mean not only “image” but physical “imprint” in matter. There was no distinction between typos and symeion (“symbol”) – indeed this contributed to the undoing of iconoclasm, for it could not be upheld that the cross functioned merely as a sign (symeion) and not as an icon (typos). In fact, the material instantiation of the cross transformed it into typos, because for the Byzantines, contrary to Walter Benjamin’s theory, it was the very process of reproduction, of being “imprinted” on matter, that invested an image with what Benjamin called “aura”, or what the Byzantines called skhesis. Skhesis has been defined as “the indestructible relationship between prototype and copy”. This relationship is based on shared form: through the process of reproduction, the icon is no longer merely an image, but a direct physical imprint of the prototype, so that model and copy can be said to share the same form. This line of reasoning sidestepped the threat of the simulacrum and allowed the Church to reinstate iconoduly as an institutionalized practice. The practice of troglodytism, however, approached the materiality of the typos in another way, one which resisted subjection to the modes of representation and authority.
Formlessness, or the inversion of the Holy Mountain
The term l’informe (“formless”) was introduced by Georges Bataille in 1929 and expanded in the 1990s by Yve-Alain Bois and Rosalind Krauss into a critical tool that could construct an alternative modernity. This “other” modern art was concerned not with the rationalism and transparent functionalism of the High Modern, but with the processes of formlessness, such as “base materialism”, “horizontality”, and “entropy”. In a sense troglodytism could be read in terms of l’informe as an alternative Byzantine culture that sought local self-determination by exploiting the slippages between form and content in the official doctrine. They found one such opportunity in the operational materiality of the typos.
Bataille said of Manet’s Olympia that it “is the negation of … mythical Olympus”, because it did not observe the formal and ideological codes of the nude. Similarly, Byzantine rock-cut chapels were the negation of the sacred mountains evoked by the cross-in-square church through their refusal of the formal codes that regulated such structures. The external aspect of the Middle Byzantine cross-in-square church, with its gradually rising profile surmounted at the center by a dome, has been read as a mnemonic reference to the holy mountains that marked the spiritual landscape of the Mediterranean. Mountains such as Mt.Sinai and Mt.Athos were considered ladders that connected earth and heaven, and through their evocation in architecture the spiritual immanence of the mountain was transferred to the church building. Cave chapels inverted this dynamic. The outward aspect of a cave chapel didn’t look like a mountain, it was a mountain, or rather the dirt and rock and physical substance of mountains. Without form, we are left with matter. To put this in terms of the informe, without content, what remains is “base materiality”.
This is even more emphatic in the interior of cave churches. Lighting, which has a well-documented symbolic and spatial function in Byzantine architecture, is reversed or nonexistent. While a cross-in-square church evoked the natural world through its exterior, the interior was the reserved as the realm of the metaphysical. Here, light, space and ornament were deployed with the ultimate aim of “denying the physicality of form”. Masonry churches were diffusely lit from above, light refracting over gilded mosaic surfaces and dissolving the materiality of the structure. The interior of a rock-cut chapel achieved exactly the opposite. If earthly matter is evident on the outside, it is even more so within the dark, dusty, and irregularly shaped cavities of the troglodyte chapels. Light inevitably comes mostly from the entrance aperture, while dome and apse remain plunged in shadows. There is no transcendence of the physical. “Base materialism”, according to Bois and Krauss, is what resists idealism, is “whatever does not allow itself to be in-formed”, whatever reason cannot cover with what Bataille called a “mathematical frock-coat”. The irregularity and non-measureability of the surfaces evades the symbolic interpretation of church geometry. In Pantalica, the reuse of tombs was another manifestation of l’informe. The subsumption of chronology and program in matter averts any pretense of idealism. So does entropy, another aspect of formlessness – even when in use, the atmosphere inside cave dwellings is clogged with dust, and erosion always threatens to erase them entirely.
Why did the troglodytes embrace the architecture of formlessness? Given the pervasive symbolic codification of the Middle Byzantine church, it is difficult to imagine that the carvers were not aware of what they were denying by choosing to hollow a church out of rock. There is no reason to suppose that the monks and villagers were dissident atheists acting out against religion. Part of their motivation may have been a will to local self-determination and identity that manifested in the resistance of institutionalized form. Perhaps another clue can be found in alternative attitudes towards icon worship that were not fully eradicated in the provinces by post-Iconoclastic doctrine, and a concept of the simulacrum that placed greater emphasis on the material imprint of the divine, rather than on the process of reproduction that posited an “indestructible” link between copy and prototype.
The simulacrum and l’informe are operations that renegotiate the relationship between image and material embodiment. Both undermine conventional assumptions of technological causality and the stability of boundaries between form and content. Byzantine cave architecture, in response to motivations that remain obscure, engaged in this renegotiation in direct opposition to the prevailing architectural practices of the day. Rock-cut architecture bears the traces of what Katherine Hayles called “simulated nature and natural simulation” – the constantly-shifting exchange between observer and world. However, it must be recognized that troglodytism was widespread in Byzantium and also signaled participation in a wider cultural identity. Whatever aims of provincial self-assertion the cave settlements represent, they were equally outward-looking and cosmopolitan. The deeper social and ideological currents that underlie the phenomenon of Byzantine troglodytism and its paradoxes remain to be uncovered.
Ann Wharton Epstein, “The Problem of Provincialism: Byzantine Monasteries in Cappadocia and Monks in South Italy,” in Journal of the Warburg and Courtauld Institutes 42 (January 1, 1979), 28–46.
Robert Leighton, “Pantalica (Sicily) from the Late Bronze Age to the Middle Ages: A New Survey and Interpretation of the Rock-Cut Monuments,” in American Journal of Archaeology 115, no. 3 (July 1, 2011), 447–464.
Emma Blake, “The Familiar Honeycomb: Byzantine Era Reuse of Sicily’s Prehistoric Rock-Cut Tombs,” in Archaeologies of Memory, edited by Ruth M. Van Dyke and Susan E. Alcock (Oxford: Blackwell, 2003), 216.
Veronica Kalas, “Early explorations of Cappadocia and the monastic myth”, in Byzantine and Modern Greek Studies 28 (2004), 101-199.
This observation was made by the Moslem geographer Ibn Hawqal, who visited Sicily in the 970s. See Blake, The Familiar Honeycomb, 214.
See Epstein, “The Problem of Provincialism”.
Vita di Sant’Elia il Giovane, ed. G. Rossi Taibbi, Instituto Siciliano di Studi Bizantini e Neoellenici, Test e monumenti, vii (Palermo: 1962), pp. 40ff.
Epstein, “The Problem of Provincialism”, 44.
See N. Katherine Hayles, How We Became Posthuman (Chicago: University of Chicago Press, 1999),17.
Pancarlık Kilise, Ürgüp
Spiro Kostof, Caves of God: Cappadocia and its Churches (Oxford: Oxford University Press, 1989), 91.
The inscription is slightly damaged and the reading of some of the words has been contested. This translation reads phoron as phobon (fear). Scholars debate whether typos refers to “image” or is a liturgical reference to the bread and body of Christ. Also, the last typos is sometimes read as topos (place). For a discussion of the inscription and its interpretations, see Catherine Jolivet-Lévy, Les Églises Byzantines De Cappadoce: Le Programme Iconographique De L’abside Et De Ses Abords (Paris: Editions du Centre National de la Recherche Scientifique, 1991), pp.221-222.
“Prototype”would be prototypos in Greek , as opposed to typos (“image” or “imprint”).
Jean Baudrillard, Simulacra and Simulations, trans. Sheila Faria Glaser (Ann Arbor: University of Michigan Press, 1994),5.
Gilles Deleuze, “Plato and the Simulacrum.” October, no. 27 (Winter 1983), 53.
This taming of iconoduly might be seen to persist to the present day in the conventional mistranslation of the inscription in Pancarlık Kilise.
Epstein, The Problem of Provincialism, 35.
In direct reference to the incarnation of God in Christ. See Bissera Pentcheva, “Painting or Relief: The Ideal Icon in Iconophile Writing in Byzantium,” in Zograf, no. 31 (2007), 7–14.
Yve-Alain Bois and Rosalind Krauss, L’Informe. Mode d’Emploi (Paris: Centre Georges Pompidou, 1996).
Sharon Gerstel, “The Church Building in Medieval Byzantium: Rite and Passage.” In Cambridge History of Religious Architecture (forthcoming).
Slobodan Ćurčić, “Architecture as Icon.” In Architecture as Icon: Perception and Representation of Architecture in Byzantine Art, edited by Slobodan Ćurčić and Evangelia Hadjitryphonos (New Haven and London: Princeton University Art Museum, 2010), 22.
Bois and Krauss, I’Informe, 29-31.
See N. Katherine Hayles, “Simulated Nature and Natural Simulations: Rethinking the Relation Between the Beholder and the World,” in Uncommon Ground: Rethinking the Human Place in Nature. Ed. William Cronon (New York and London: W.W. Norton & Company, 1996)
Blake, The Familiar Honeycomb, 218.