Philip L. Tite. Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain

Transcription

1 Philip L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain The treatment of material culture, specifically artistic images such as we find with mosaics, tends to focus upon descriptive qualities, perceiving only the need of identification of artistic representations along with production processes. While such work is certainly essential for a clear understanding of past cultures, I wish to suggest that a more theoretical appreciation of narrativity can offer us new insights into cultural processes through material culture by means of reading archaeological remains through a textual lens. Indeed, we can approach aesthetic artefactual remains such as villa mosaics much like a grammar: individual images analogously function as words that are given semantic meaning through the syntax of the arrangement of those words through spatial juxtapositions. A distinction between what Somers and Gibson have labeled representational and ontological narrativity may help direct such an effort. Somers and Gibson note that social scientists who engage narrative as a theoretical framework tend to either follow what could be called representational narrativity or ontological narrativity. Representational narratives constitute, modes of representing knowledge (telling historical stories) [that are] representational forms imposed by historians on the chaos of lived experience 1. To engage in a representational narrative analysis is to draw upon, for instance, historical texts or archaeological data to infer the historical processes, essences, and causal development of human groups. This is especially true when we consider the metanarrative level of scholarly stories of what constitutes history and historical knowledge. Ontological narrative, however, moves beyond the text or artefact as simply a repository of historically knowledge, and approaches a text as a discursive fragment wherein social agency and social structure are dialogically interrelated within an ongoing process of defining, determining, contesting and even ignoring ontological products (i.e., the self both individually and interpersonally) 2. Due to the centrality of relationality, ontological narrativity places stress upon the generation of meaning or identity through interactive processes between social actors. Such ontology is not simply reflected by narratives, but rather it is through narratives that social identities 1 SOMERS, GIBSON 1994, See SOMERS, GIBSON 1994,

2 P. L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain Abb. 1 Frampton mosaic (reproduced with permission from COSH, NEAL 2005, 135). are constituted and thus rendered meaningful. Consequently, by looking at not simply the descriptive qualities of mosaics (a form of representational narrativity) but at the discursive qualities underlying the selection, placement, and transformation of those images, we may gain a better appreciation for the interaction of the constellation of images and the viewer-as-reader. Thus, we can read the archaeological data as a discursive moment or set of moments. Such an approach places stress upon the active rather than passive role of the viewer3, recognizing that meaning is discursively established4. Today I would like to apply such a reading strategy to the villa mosaics discovered near Frampton in Dorset, the most spectacular being mosaic (following Cosh and Neal s numeration)5. These mosaics are of particular interest to me, being a scholar specialized in the study of ancient Gnosticism and given the recent arguments over the possible presence of 3 See GIOVANNELLI Cf. RICOEUR 1991, COSH, NEAL 2005, Bollettino di Archeologia on line I 2010/ Volume speciale E / E10 / 4 Reg. Tribunale Roma n. 330 ISSN

3 XVII International Congress of Classical Archaeology, Roma Sept Session: Roman and Barbarian Gnosticism within fourth-century Roman Britain. The site was discovered in 1794 by labourers and presented by James Engleheart to the Society of Antiquaries of London in A fresh excavation by Samuel Lysons in resulted in a spectacular lithographic reproduction of the mosaics along with his report on the mosaics 6. Unfortunately, the mosaics were destroyed, supposedly in the 1850s due to troop movements, and we are left with Lysons drawings for our analysis of these mosaics 7. Mosaic is a tripartite room, comprised of an antechamber, a main room, and an apse (fig. 1). The antechamber is comprised of a central image that has been identified as Bacchus (due to the remnants of a rider on a leopard and what looks like a thyrsus) 8, flanked on both sides by typical hunt scenes (with a lion and hunter at the threshold and, opposite, a deer being hunted by a hunter and dog). The main room has a central image of what most scholars have seen as Bellerophon, on Pegasus, slaying the Chimaera 9. Around this image are four square panels that seem to include mythological pairs, the identification of which is contested (e.g., pairs of lovers, Venus and Adonis, Paris and Oenone, Attis and Sagaritis, or the judgment of Paris) 10. We also have four semi-circular panels, which have all been destroyed except the fragments of what looks like a swan s head, a T shape remnant, and the folds of a garment or rocks. Surrounding the entire room is a border of opposing pairs of dolphins, emerging from the beard of Neptune, who faces the entrance to the apse. These pairs of dolphins metamorphically change into some form of bird (likely herons or swans) 11 when they reach Cupid at the other entrance to the room. There are two inscriptions at each entry point, clearly identifying Cupid and Neptune. The apse contains a central image of a cantharus, badly damaged and surrounded by various geometrical shapes most notably the eight-strand guilloche (which draws attention to the cantharus,). Between the main room and the cantharus, is a set of seven spherical patterns, the central one being an obvious chi-rho. Dominic Perring reads this tripartite room as a Gnostic narrative of the ascent of the soul. He offers the most sophisticated and fully developed argument for identifying Gnosticism in late Roman Britain, specifically among the non-urban elite whose magnificent villas flourished in the fourth century 12. His work demonstrates an appreciation for the discursive qualities of the Frampton mosaics, arguing that rather than a random selection of images, the very organization moves the reader through a mythical narrative. There are several elements that support a possible Orphic- Bacchic Christian dualism as a subtext for these mosaics; a dualism that Perring places under the label Gnosticism. With the images of the mythological pairs and Bellerophon slaying the Chimaera, a sharp motif of dualistic struggle emerges. Added to the duality is the ascent of the soul motif, so common within various Gnostic systems, reflected in the iconographic utilization of the dolphins. The Christian monogram reshapes this entire constellation of images, giving it a clearly Christian nuance. Perring s reading is centered upon following a plot development through the three rooms: the antechamber moving the reader into a state of soul or initiation and the central chamber establishing a cosmological conflict of matter placed within a series of dualistic pairs, while the apse establishes a sacramental synthesis of soul and matter leading toward 6 LYSONS 1813; COSH, NEAL 2005, The trustworthiness of Lysons s pictures are generally accepted; see, for example, JOHNS LYSONS 1813, 4; PERRING 2003b, 105 6; COSH, NEAL 2005, COSH, NEAL 2005, 135. This identification is based upon a comparison with the villa mosaics at Lullingstone and Hinton St. Mary; contra LYSONS 1813, 3, who understood this as a man on horseback, combating a lioness. 10 See discussion in COSH, NEAL 2005, , as well as the sources given on proponents of the various readings. 11 COSH, NEAL 2005, 136, who credit this observation to Patricia Witts. 12 PERRING 2003a; 2003b. 43

4 P. L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain Abb. 2 Re-directed perspective at chi rho: discursive positioning of reader as aesthetic narrative participant (non-passive engagement with text) (modified from C OSH, NEAL 2005, 135). spiritual unity or harmony (through the cantharus and a privilege position of viewing the rooms from the chi-rho). Furthermore, a three-point perspective is given, where the reader actively shifts her or his perspective at each step of the journey to the apse (fig. 2). The shifting perspective renders the viewer a narrative participant, taking a discursive role within the very narrative procession. For Perring, such progression of the viewer reflects a process of initiation, though he stops short of declaring Frampton a cultic centre13. There are three advantages to Perring s narrative approach to Frampton. First, he reads the mosaics as narrative, linking the referentiality of this text to the broader cultural and intellectual climate of late Roman Britain. This is a significant move away from simply identifying iconographic images, as if such identification were the end goal of art history. Second, he considers the images in connection with other images, rather than isolating a panel or set of images (previous studies of Romano-British mosaics tend to fall into this problem, such as interpretations of the cockerelheaded figure at the Brading villa). Rather, Perring looks for patterns within the organization, juxtaposition, and progression of the images. Thus, he renders the Frampton mosaics narrative texts that can be read or re-actualized by the act of (re-)reading. Finally, Perring is the only 13 Cf. HENIG 1984, 173. Bollettino di Archeologia on line I 2010/ Volume speciale E / E10 / 4 Reg. Tribunale Roma n. 330 ISSN

5 XVII International Congress of Classical Archaeology, Roma Sept Session: Roman and Barbarian Romano-British scholar, to my knowledge, to engage the emerging scholarship on Nag Hammadi and Gnosticism. However, there are also a few shortcomings in Perring s reading of Frampton. First, he largely depends upon English translations of the Nag Hammadi tractates, while simultaneously failing to distinguish them geographically or by sectarian leanings. Second, his heavy reliance upon the Church Fathers, most notably Hippolytus, for appreciating Gnostic syncretism, overlooks the rhetorical strategies taken by the Fathers to polemically discredit their opponents (e.g., Hippolytus establishes a polemical view of Greco-Roman philosophical and mythical traditions and then locates the error of the heretics within that non-christian setting). Thirdly, a major problem with not only Perring s treatment of Gnosticism, but also of the entire discussion of Gnosticism in Roman Britain is the treatment of the very category Gnosticism. In recent scholarship the selfevident, transhistorical view of this category has been challenged, prompting several to discard the category as problematic and ideologically driven 14 despite continued attempts to discern an original Gnostic religion or innovation 15. Fourth, Perring may be forcing his text to fit a Gnostic system (i.e., the matter-soul-spirit progression) and Gnostic cultic setting. While appreciating the direction taken in reading the Frampton mosaics, and seeing great value in locating Gnosticism within late Roman Britain for my own field of study, the argument set forth strikes me as extremely speculative. While some form of Christian dualism may have existed within the late Romano-British religious landscape, to identify this as part of Gnosticism will require greater theoretical discernment. Still, his work offers us a sophisticated and potentially insightful reappraisal of these mosaics. I am especially impressed with Perring s sociological application of his interpretation, linking the iconography at Frampton to a rural aristocratic resistance to emerging urban ecclesiastical authority. Using Perring s reading as a beginning point, I believe we can re-read these mosaics in a similar way that we interpret religious texts ; i.e., to treat our non-text as textual. An observation that is overlooked by Perring is Lysons s initial comments on the chi-rho monogram, an image central for Perring s own interpretation. Lysons suggested, based on artistic style, that the Christian monogram was likely a later addition 16. His comments, however, can be read in different ways at one point he seems to suggest that only the chi-rho was a later addition to the foliate scroll 17, while elsewhere that the entire apsidal extension was added later 18. Unfortunately, the destruction of these mosaics renders it impossible to test either understanding of Lysons s suggestion. Indeed, Roger Ling has recently raised a persuasive argument against seeing two periods in the construction of these mosaics. Ling s argument is largely based upon the Christian elements within the Hinton St. Mary mosaics, which were most likely produced by the same workshop as the Frampton mosaics (the Durnovarian group) (see fig. 3). Given both the ambiguity of Lysons s suggestion, the non-stratification theory set forth by Ling, and our inability to test any of this due to our dependence on Lysons s drawings, any reading of the Frampton mosaics must be treated as tentative and somewhat speculative. Still, the possibility of a later addition of the Christian elements at Frampton suggest that we may profitably read this artistic narrative through redactional 14 See WILLIAMs 1996; TITE 2001a; KINg For example, SMITH 2004; PEARSON LYSONS 1813, 3, LYSONS 1813, 3, 6; COSH 2003, A reading followed by LING 2007a, 8. 45

6 P. L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain layers for constructing a history of the mosaicas-text, teasing out potentially shifting discursive moments in this history. Redaction criticism along with stratification theory is widely used within the study of early Jewish and Christian texts, including the Nag Hammadi codices19. The stress falls upon changes in texts as indices of rereadings i.e., additions, eliminations, or relocation of material by a later editor/reader of a text. By utilizing his or her textual material, each redactor ( editor ) takes on the role of author ; specifically, by making adaptations or re-writings of a source, new meanings are infused into the text, thus producing a new text. Redactors will vary in how much of an authorial role they play, as they vary in the intrusive engagement with their source material20. The redaction critic strives to uncover seams that are indicative of such editorial activities. As redactional activity is closely tied to the redactor s social situation, social context, and ideological agenda, an analysis of the editorial hand should give the historian a glimpse into the community setting or occasion of writing. In discerning these shifts Abb. 3 Hinton St Mary mosaic (reproduced with through establishing compositional layers, the permission from COSH, NEAL 2005, 158). historian is led to reconstructing the compositional history of a text, and thus the underlying community of a text, through diverse strata. Although this approach to texts is largely inferential in the historical products it produces (i.e., rendering texts simply as documentary windows into the life setting behind a text), an appreciation for compositional histories need not fall into a type of representational narrativity. Rather, we can view redactional activity as keys to unlocking the discursive moments within the history of that text. Paul Ricoeur summarized the discursive function of the text: If reading is possible, it is indeed because the text is not closed in on itself but opens out onto other things. To read is, on any hypothesis, to conjoin a new discourse to the discourse of the text. This conjunction of discourses reveals, in the very constitution of the text, an original capacity for renewal which is its open character. Interpretation is the concrete outcome of conjunction and renewal 21. Consequently, each redactional stage becomes a moment when a text is opened or re-opened through the act of re-reading. New discourses, along with their attendant meanings, are created through the conjoining of the text within a new context. By isolating points of such conjunction, we may gain new insights into re-presentation of stories as 19 Helpful overviews of redaction criticism are offered in BRIDGE 2002; PERRIN For a discussion of redactor as author, see TITE 2001b. 21 RICOEUR 1991, Bollettino di Archeologia on line I 2010/ Volume speciale E / E10 / 4 Reg. Tribunale Roma n. 330 ISSN

7 XVII International Congress of Classical Archaeology, Roma Sept Session: Roman and Barbarian they bear upon and guide the construction of new ontological perceptions and identities of social life. Not only is such an approach valid for written texts, but also, I contend, for those narratives that are embedded within material culture. There are three possible redactional stages discernable with the Frampton mosaics. Each stage works with the theme of tolerance or intolerance. The indications for such stages include the possible addition to the mosaic (the Christian monogram or even the apse in its entirety) and the pre-excavation destruction of the mosaics. Due to space constraints, I will only offer one possible line of development, specifically the possibility that the foliate scroll may have been re-laid with a chi-rho facing the image of Neptune. The first redactional stage is one of religious tolerance for diverse mythical and moral motifs. Upon entering this tripartite room, the viewer is immediately faced with two paralleled hunt scenes in the antechamber. Hunt scenes carried a virtue motif, indicating courage or manliness. 22 This virtue motif sets the tone for the Orphic-Bacchic themes that dominate the entire narrative journey. Indeed, the Orphic-Bacchic motifs are arranged within a narrative progression from Bacchus (flanked by a virtue motif) to the cantharus in the apse. This opening-and-closing function of Bacchus is certain when we recognize, in Cosh s words, that the cantharus was not only symbolic of wine, but also of Bacchus 23. Thus, the viewer begins with Bacchus (antechamber) and ends with Bacchus (apse). By using Orphic-Bacchic images to open and close the narrative, the very innertexture 24 moves the viewer from the call to virtue to the attainment of the mysteries through freedom from this dualistic conflict set forth in the main room. Along with images taken from the mysteries and typical hunt scenes, stage one also includes various Greco-Roman mythological motifs (Cupid, Neptune, Bellerophon battling the Chimaera, and the pairs). We may also have astrological concepts worked into the narrative: the seven spheres leading into the apse, representing the seven heavenly spheres through which the soul must journey; which is reinforced by the metamorphic motifs of the dolphins transforming into water fowl (dolphins, water fowl, and especially metamorphic and polymorphic figures tend to carry liminal qualities, usually between the realms of life and death or material/corruptible and spiritual/incorruptible realms; it is also possible that the T shape remnant may be the distorted tail of a dolphin undergoing a transformation, though this interpretation of the fragment is very speculative). By drawing upon diverse images, our author has set forth a journey motif through the narrative opening, middle and closing. While Perring connects this room-by-room progression as soul-matter-spirit, I would suggest that instead of cultic initiation we may have a narrative more closely aligned with asceticism. The framing of this journey by Orphic-Bacchic motifs sets this more in line with the mystery religions than Gnosticism (at least at stage one). As set forth in Figure 4, the antechamber s call to virtue or courage draws the viewer into a station of initiation (µύ σις; myēsis). This initial station on the journey into the mysteries functioned to prepare the initiate through purification. If Simms is correct, then within the Eleusinian mysteries, µύ σις may have functioned as the overarching concept for the subsequent stations of perfecting (τeλeτή; teletē) and vision (e ποπeία; epopteia) 25. Such an overarching function may fit the narrative progression of the Frampton mosaics, where a call to virtue connected to Bacchus may set forth the ascent of the soul that dominates the main room. Consequently, the main room (as the τeλeτή) 22 See, e.g., TUCK COSH 2003, See the socio-rhetorical criticism proposed by ROBBINS SIMMS

8 P. L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain Abb. 4 Perring s identified dualistic structures, overlaid with modified narrative progression: asceticism rather than cultic initiation (modified and used with permission from PERRING 2003a). presents the ascetic struggle that is largely a renouncement of the world, thematically supported by the metamorphic transformations (liminal movement between realms; not uncommon in ascetic texts), dualistic pairs (likely connected to life and death; which effectively elucidates the virtue motif of the antechamber), and the central mythical struggle of Bellerophon. Unlike the Eleusinian mysteries, the mysteries of Dionysus were, as Joscelyn Godwin aptly notes, initiations through actions rather than beholding 26. The main room at Frampton effectively demonstrates such action-driven progression, in this case through means of ascetic ascent from humankind s titanic condition. Finally, with the cantharus, the viewer reaches the climax of this ascetic struggle; i.e., 26 GODWIN 1981,

9 XVII International Congress of Classical Archaeology, Roma Sept Session: Roman and Barbarian the mystical unification of e ποπeία;. Having reached this final point of vision of seeing the god or seeing from a divine perspective, it is not surprising that the viewer s perspective is reoriented towards a new view of the pathway just followed (the climax of the three-point perspective). In effect, the viewer has become a Bacchos, with a completely new view on reality 27. The choice of the cantharus, as a symbol of wine and Bacchus, makes a great deal of sense for this narrative closing given Bacchus s role as liberator of inhibitions 28. With the achievement of becoming a Bacchos, the viewer is now able to descend back into the world. It is perhaps this descent motif that is established through the spatial perspective granted by the apse. Indeed, the two inscriptions can only be read if a person were to turn around and face them or as they leave the room 29. The diverse images drawn upon to play out this narrative progression need not suggest that the cult of Bacchus was present in Frampton, or any other Romano-British villa where such iconography is present. Rather, we may simply have a decorative function; a decorative choice that would likely have suited a public or feasting use of this tripartite room (if that was one of its uses) 30. This decoration, however, likely carried a narrative function; telling a story that weaves virtue, mysteries, struggles, myth, and attainment. By playing with images as well as space, the author effectively draws the viewer into the story, to become an active reader who plays a role in the narrative literally step-by-step. In constructing such a creative narrative, the author playfully weaves diverse, though typical, motifs together. In the second redactional stage there is a Christianization of this Orphic-Bacchic narrative. This is the stage at which the chi-rho is added to the foliate scroll, and it is this seemingly minor element that facilitates a fresh re-reading of Frampton. As Ling has argued at length, the other elements in this set of mosaics need carry no Christian connotations: the Bellerophon-Chimaera panel is not specifically Christian (despite the general sense of battling evil) 31 ; the inscriptions carry no Christian sense, but rather points toward the deities within the main room (Cupid and Neptune); and the rest of the images are typical of Roman mosaics 32. Ling correctly observes that, Were it not for the presence of the chi-rho symbol, the Bellerophon pavement at Frampton would not have provoked much debate 33. I fully agree with him on this point, and would extend his comments to the cantharus as an Orphic-Bacchic symbol. However, we must contend with the fact that the Christian monogram is present and that this addition infuses the entire narrative with a Christian nuance 34. Unlike Perring and Toynbee, however, I would not see this Christian element as necessarily part of the original design 35. From a redactional analysis, this Christian element offers several insights into the re-reading of this narrative. By placing the chi-rho into the middle of the seven spheres, with it facing Neptune rather than the cantharus, a strong Christian element is added to the entire astrological quality of the ascent motif; i.e., this is not just any journey of the 27 As TURCAN 1996, 295, succinctly puts it, What was the ultimate aim of Dionysism? To become a Bacchos, that is to say, to become identified with the god. 28 See COSH 2003, I wish to thank Colleen Cummings for first drawing my attention to this positional aspect of the inscriptions. On the translation problems associated with these inscriptions, see the valuable study in LING 2007a; 2007b. 30 COSH 2003, Contra HENIG 1984, LING 2007a, LING 2007a, Contra LING 2007a, 9, who contends that there is no Christianization of these elements. In part, Ling s argument is designed to argue against a programmatic design at Frampton, preferring to see these elements as conventional repertoire in Romano-British villa decoration. 35 TOYNBEE 1968; PERRING 2003b; 2003a. 49

10 P. L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain soul, but a journey that climaxes with a Christ motif as redeemer figure. Secondly, it is fascinating that our redactor does not reject the non-christian elements. This is not an instance of ideological contestation or polemic against a broader religious and cultural landscape, such as we find with Hippolytus Refutatio omnium haeresium or the Valentinian Gospel of Philip (especially 72,1-5). Rather, the redactor of the Frampton mosaics playfully draws upon that broader culture, seeing its narrative as part of its own Christian narrative. As with the first redactional stage, we see here a high degree of tolerance. Christian and non-christian religious imagery was evidently not a problem for this redactor. Such tolerance suggests not only educated elites using this villa, but also a high cultural accommodation for constructing a religious identity (not all that dissimilar to the Simonian Gnostic incorporation of Homeric narrative to shape the mythic account of Helena or the soul s descent/entrapment see Irenaeus, Adv.haer ; cf. Exegesis on the Soul, which is likely Simonian). Thirdly, by appropriating the existing motifs, these images are discursively viewed through a playful Christian re-reading, enabling readers to view these motifs as part of a Christian narrative of the dualistic struggle and ascent to Christ. Simultaneously, however, a reader can still view these motifs through the previous interpretative lens. We see this coming out profoundly with the cantharus, which was identified as part of Orphic-Bacchic motifs framing the narrative. With the chi-rho, however, this cup can now be re-perceived as a Eucharistic cup that complements not Bacchus in the antechamber, but rather the Christ motif at the centre of the astrological threshold. Indeed, just as the cup stood as a symbol of Bacchus in the first redactional stage, it now is a symbol not simply for but of Christ. Alternatively, the cantharus could still be read as a parallel to the antechamber in our narrative progression of opening-middle-end. It is this playful aha! that gives the narrative its intellectual power; i.e., it draws the viewer even more into the narrative, forcing the viewer to draw inferences from the images as if there was a pun worked into the pictorial syntax. Whether the viewer as reader comes to the same conclusion as other readers is part of the playfulness. Both Perring and Ling indicate that the Frampton mosaics likely were designed to evoke discussion between hosts and guests 36. Such discussion would demonstrate the cultural values of otium or paideia. The aha! moment is when one demonstrates, and subsequently recognizes in the host, a cultivated mind. The utilization of Dionysiac/Bacchic imagery within a villa context, in particular, was not just an attractive decoration or an appropriate theme for a dining room but also to indicate their knowledge of and adherence to classical culture 37. This realignment of the images syntactical relationship alters the progression of the narrative. Rather than a fulfillment of the promise of virtue-in/through-bacchus (the opening and closing acting as framing mechanisms), the viewer is guided towards a Christological climax. Indeed, if Bacchus is replaced by Christ in the apse, then perhaps Bacchus in the antechamber, upon entering the apse, becomes a polymorphic manifestation of Christ 38. A re-perception of the motifs comes into play upon reaching the apse and one is given a new, perhaps deeper understanding of what he or she has just seen or read. The ascetic struggle in the main room certainly fits with this realignment, while simultaneously not negating the original narrative components from stage one. Within early Christian narratives, the ascetic s journey or the visionary s ascent is closely tied to elements of dualistic struggle against life and death, 36 PERRING 2003b, 97; LING 2007a. 37 WITTS 2001, 17. See also MILLETT 2005, 98; 1992, For an excellent discussion of polymorphic and metamorphic Christology, see FOSTER

11 XVII International Congress of Classical Archaeology, Roma Sept Session: Roman and Barbarian metamorphic and polymorphic manifestation (calling into question the realness of the earthly realm), ascent through heavenly spheres, and liminal danger between life and death. Underlying such narrative elements is a strong call to virtue and purification. Within the apocryphal acts accounts we find especially polymorphic and ascetic qualities brought together in opposition to the world s vices and entanglements (especially marriage and procreation) (e.g., Acts of John and Acts of Thomas). In the Acts of Peter and the Twelve Apostles we find perhaps the best textual parallel to Frampton s narrative progression. In this account, the apostles are directed by Lithargoel, a pearl merchant, to travel from the island city of Habitation to the merchant s city Nine Gates. The apostles are commanded to take nothing with them, due to the dangers facing them along the way (robbers, wolves, lions, and bulls). The journey oddly takes place on water, despite the land based travel dangers in the exhortation to renounce the world. When the apostles survive the sea storms and arrive at Nine Gates (certainly carrying astrological connotations), they meet with Lithargoel, who takes the shape of a physician who then reveals himself as the Saviour. The unveiling results in illumination of the apostles, who are then commissioned to return to the world to heal bodies and souls. We see similar elements in the ascent motif at Frampton: a preliminary encounter with a divine being who calls the viewer to virtue (Bacchus and the hunt scenes); the dangerous struggle to overcome this world (Bellerophon and the Chimaera); a maritime motif infused with a journey of the soul (Neptune and the dolphins); an astrological ascent (not only Pegasus but more importantly the foliate scroll of seven spheres); and a metamorphic transformation of not only the one on a journey (dolphins into water fowl), but also of the divine being. Indeed, with the Christian redactional element, Bacchus is revealed fully as Christ as one enters the apse (chi-rho and cantharus, with the cantharus drawing one back to a reassessment of the Bacchus figure in the antechamber). Such full revelation of Bacchus-as-Christ is not dissimilar from Lithargoel s progressive revelation as the Saviour. The dualistic qualities in the main room are essential for understanding the struggling aspect of those panels; though with the Christian element read onto them, they could take on a Gnostic quality but such identification is not necessary for appreciating the Christian dualistic re-reading being played out. Of course asceticism and spiritual ascent is not divorced from Gnosticism. Many would, for instance, read the Acts of John and the Acts of Thomas (or sections at least) as Gnostic, and even the Acts of Peter and the Twelve Apostles is heavily debated with regard to its connection to Gnosticism. One of the key problems, however, is that Gnosticism is not a clearly defined category for a self-evident group in antiquity but rather is a modern construct for understanding (or perhaps misunderstanding) diverse cultural and religious phenomena. Still, Perring may have identified a form of Christian dualism in fourth-century Roman Britain that may be studied alongside other Christian dualistic groups that tend to fall under the label Gnosticism. Regardless of how one decides on the Gnostic connection to Frampton, what is evident is that at redactional stage two there is a Christianizing process at work following along the lines of playful tolerance of Christian and non-christian motifs. Redactional stage three differs from stages one and two in that there is an almost certain intolerance being demonstrated. This final stage is marked by the pre-excavation destruction of select panels. The destruction is almost certainly intentional rather than random 39, due to the methodical care taken in destroying them (e.g., the four semi-circular panels in the main room) while other panels are less obviously intentional (e.g., the cantharus and the Bellerophon panels). 39 So also PERRING 2003b,

12 P. L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain This selectivity may provide indices for ideological conflict or social contestation in the late fourth century. Non-Christian images destroyed include: main room: four semi-circular images; antechamber: possibly the Bacchus image (if the damage is intentional). Non-Christian images preserved include: main room and antechamber: Cupid and Neptune; likely corner pairs; hunt scenes; perhaps Bacchus image (if the damage is accidental). Christian images destroyed: apse: cantharus (= Eucharist cup). Christian images preserved: apse: the chi-rho. From this list, it is surprising that the destruction and preservation is not focused exclusively on non-christian or Christian images. Rather, the final redactor has intentionally destroyed several images that were evidently offensive, yet left other images intact; thus, the redactional activity is not an anti-christian nor a pro-christian act of intolerance, but some odd mixture. Of course, as Ben Croxford has correctly and insightfully demonstrated, the damage inflicted upon Romano- British images need not be acts of hostility, but rather could be acts of appropriation and reutilization (giving the selected pieces no longer present a new use or meaningfulness due to a positive valuation; thus what remains for the archaeologist need not be what was valued but rather what was discarded) 40. The damage at Frampton, however, is more likely a hostile act due to the incompleteness of the destruction of some panels. For example, the central panel in the main room is neatly destroyed around the edges but the main image of Bellerophon slaying the Chimaera is not completely removed, thereby suggesting that this panel s destruction was intentional but not removed for use elsewhere (so also with the cantharus). Why were these panels destroyed? There are three possibilities. First, as Perring argues, a religious or socio-economic conflict between rural aristocracy and urban ecclesiastical authority may underlie this final stage at Frampton. This explanation places stage three into an orthodoxy vs. heresy model, with the destruction linked to breaking the iconography s magical power. While this explanation is possible, it is fraught by a problematic historical model that many of us see as ideologically driven. More importantly, it does not account for the selective process of what is destroyed. Why preserve the chi-rho and destroy the cantharus, if both are standard Christian symbols, while keeping Neptune as a clearly non-christian image? A second possible explanation is an internal conflict between Christians over Eucharistic worship. Specifically, I have toyed with the idea that the destruction could have been done by Pelagian Christians who may have accepted the chi-rho but contested the sacramental aspect of the cup. The problem with this explanation is that it also does not account for the non-christian elements preserved or destroyed. A third possibility, which I think far more likely, is an attack on esoteric approaches to Christian worship. While many Christians saw no conflict of interest with classic mythological images drawn from Greece and Rome (seeing them as stories in need of allegorizing; not unlike the treatment of fairy tales today), they may have found an esoteric means of grace a threat to orthopraxy. Such a rereading of the mosaics would necessitate a new narrative interpretation of the images, specifically the cantharus. In this re-reading, the cantharus is no longer seen through a Christian lens, but rather is linked to the Orphic-Bacchic mysteries (as it was in stage one). At redactional stage three, the redactor may have been focused upon negating esoteric mysticism to the extent that other 40 CROXFORD 2003, especially

13 XVII International Congress of Classical Archaeology, Roma Sept Session: Roman and Barbarian elements, most notably the astrological and metamorphic motifs, may have been overlooked. Striking from the perspective of ontological narrativity, is that while stages one and two shape or re-shape their narratives to encourage the viewer to become an active reader of the images, at stage three the redactor s role of author is founded upon taking up the role of reader and a concern to negate the possibility of others reading the offensive narrative of the mosaics. Thus, at stage three we are faced with redactional censorship, which, through the fresh conjunction of discourses, ironically produces a new text. There has been a growing interest in interdisciplinary cooperation between archaeologists and ancient historians, bringing together material and textual evidence for our reconstructions of the past. Indeed, Zsuzsanna Várhelyi sees such cooperation as the greatest opportunity for both classical archaeology and ancient social history, while recognizing that there are numerous potential challenges that we ought to consider seriously in order to benefit fully from this collaboration 41. This paper attempts to contribute to this trend; encouraging those of us who study ancient religious traditions (textually or artefactually) to more fully appreciate the discursive, identity-constructing role of religious iconography. The Frampton mosaics offer an ideal test ground. Here we have potential redactional layers within our text that serve as indices of changes in narrative meaning: each reactivation of this text is evidence of a fresh re-reading. I have suggested that we can read Romano-British mosaics by exploring the grammatical dynamics of space and images as syntax giving words meaning. Such an interpretative move shifts the analytical focus from either using material culture to illuminate texts or texts to elucidate material culture 42, to a collapsing of the disciplinary divides; i.e., material remains function as texts in that they are an engaged element within discursive processes of establishing, contesting, legitimating, or redefining social identity. Acknowledgements I wish to express my appreciation to Colleen Cummings, Dominic Perring, and Michael Williams for their helpful comments during the preparation of this paper. I also wish to thank Dominic Perring, Stephen Cosh and David Neal for kindly granting permissions in preparing figs Philip L. Tite Independent researcher, Seattle WA USA philip.tite@mail.mcgill.ca 41 VÁRHELYI 2006, 313; cf. STOREY 1999; GARDNER See the discussion in DYSON 1994, especially on the impact of post-processual theories on the relationship of text to archaeology; cf. FALKENHAUSEN

14 P. L. Tite Reading and Re-Reading the Frampton Mosaics: Religious Innovation and the Construction of Cultural Identity in Roman Britain Bibliography BRIDGE S. L., Literary Source and Redaction Criticism. In A. J. BLASI, J. DUHAIME, P.-A. TURCOTTE (eds), Handbook of Early Christianity: Social Science Approaches. New York, COSH S. R., Mosaics and Wine. Mosaic, 30, COSH S. R., NEAL D. S., Roman Mosaics of Britain. Volume 2: Southwest Britain. London. CROXFORD B., Iconoclasm in Roman Britain? Britannia, 34, DYSON S. L., Is There a Text in this Site? In Methods in the Mediterranean: Historical & Archaeological Views on Texts & Archaeology. Leiden, FALKENHAUSEN L., On the historiographical Orientation of Chinese Archaeology. Antiquity, 67, FOSTER P., Polymorphic Christology: its Origins and Development in Early Christianity. Journal of Theological Studies, 58(1), GARDNER A., Identities in the late Roman Army: material and textual perspectives. In G. DAVIES, A. GARDNER, K. LOCKYEAR (eds), TRAC 2000: Proceedings of the 10th Annual Theoretical Roman Archaeology Conference. Oxford, GIOVANNELLI A., In and Out: the Dynamics of Imagination in the Engagement with Narrative. Journal of Aesthetics and Art Criticism, 66(1), GODWIN J., Mystery Religions in the Ancient World. San Francisco. HENIG M., Religion in Roman Britain. London. JOHNS C., Samuel Lysons: a Founding Father of Romano-British Archaeology. Mosaic, 27, KING K. L., What Is Gnosticism? Cambridge. LING R., 2007a. The Bellerophon Mosaic at Frampton. Mosaic, 34, LING R., 2007b. Inscriptions on Romano-British Mosaics and Wall-Paintings. Britannia, 38, LYSONS S., Account of a Discovery of Mosaic Pavements Near Frampton in Dorsetshire in the years 1794 and In Religuiae Britannico-Romanae. Containing Figures of Roman Antiquities Discovered in Various Parts of England. London, Vol.1, 1 6. MILLETT M., The Romanization of Britain: An Essay in Archaeological Interpretation. Cambridge. MILLETT M., Roman Britain. New York. PEARSON B. A., Ancient Gnosticism: Texts and Literature. Minneapolis. PERRIN N., What is Redaction Criticism? Eugene. PERRING D., 2003a. Deconstructing the Frampton pavements: gnostic dialectic in Roman Britain? In G. CARR, E. SWIFT, J. WEEKES (eds), TRAC 2002: Proceedings of the Twelfth Annual Theoretical Archaeology Conference, Canterbury Oxford, PERRING D., 2003b. 'Gnosticism' in Fourth-Century Britain: the Frampton Mosaics Reconsidered. Britannia, 34, RICOEUR P., What is a text? Explanation and understanding. In M. J. VALDÉS (ed), A Ricoeur Reader: Reflection and Imagination. Toronto, ROBBINS V. K., The Tapestry of Early Christian Discourse: Rhetoric, Society and Identity. London. SIMMS R. M., Myesis, Telete, and Mysteria. Greek, Roman and Byzantine Studies, 31,

15 XVII International Congress of Classical Archaeology, Roma Sept Session: Roman and Barbarian SMITH C. B., No Longer Jews: The Search for Gnostic Origins. Peabody, MA. SOMERS M. R., GIBSON G. D., Reclaiming the Epistemological Other : Narrative and the Social Construction of Identity. In C. CALHOUN (ed), Social Theory and the Politics of Identity. Oxford, STOREY G., Archaeology and Roman Society: Integrating Textual and Archaeological Data. Journal of Archaeological Research, 7 (3), TITE P. L., 2001a. Categorical Designations and Methodological Reductionism: Gnosticism as Case Study. Method and Theory in the Study of Religion, 13, TITE P. L., 2001b. Textual and Redactional Aspects of the Book of Dreams (1 Enoch 83-90). Biblical Theology Bulletin, 31(3), TOYNBEE J. M. C., Pagan Motifs and Practices in Christian Art and Ritual in Roman Britain. In M. W. BARLEY, R. P. C. HANSON (eds), Christianity in Britain, Leicester, TUCK S. L., The Origins of Roman Imperial Hunting Imagery: Domitian and the Redefinition of Virtus under the Principate. Greece & Rome 52(2), TURCAN R., The Cults of the Roman Empire. Oxford. VÁRHELYI Z., Social History and Archaeology: Opportunities and Problems. In C. C. MATTUSCH, A. A. DONOHUE, A. BRAUER (eds), Common Ground: Archaeology, Art, Science and Humanities: Proceedings of the XVIth International Congress of Classical Archaeology, Boston, August 23-26, Oxford, WILLIAMS M. A., Rethinking "Gnosticism": An Argument for Dismantling a Dubious Category. Princeton. WITTS P., Universal Messages: Iconographical Similarity Between the Mosaics of Antioch and Britain. Mosaic, 28,