Introduction

The Skenographia Project: Investigating Pompeian Wall Paintings through Virtual Reality.

The Skenographia Project was funded by the Leverhulme Trust, and took place from April 2002 to October 2003.

Despite the extensive research focussed for over two centuries upon Roman wall paintings, in certain ways they still represent a vast, inadequately exploited source of knowledge. In recent years, research in this area has increasingly turned from formalist analyses of style and influence, to the study of these paintings in their architectural settings and as part of a complex cultural interplay of signs through which aesthetic, social, cultural, and economic values were produced and negotiated (e.g. Gazda (ed.), Roman Art in the Private Sphere, 1991).

However, one important area of investigation has remained relatively neglected, namely: research from a theatre-historical perspective. On the one hand, the paintings offer a great deal of evidence for ancient theatre practice; a certain amount of work has been carried out in this area, but a great deal more needs to be done. On the other hand, little or no scholarly attention had been given to analysing how modes of spectatorship associated with the theatre may have influenced the design, display and reception of these insistently theatrical domestic paintings. In what is otherwise a vibrant and dynamic field of academic enquiry into the period in question and its cultural products, these were critical lacunae in contemporary scholarly discourse, and as such deserved urgent attention.

From a very early date, as Aristotle attests (Poetics 1449a), scenic painting—skenographia—was a constituent element of Greek theatrical performance. In the temporary stages recorded as having been built in Rome from about the 3^rd century BC well into the Imperial period, and in Roman permanent stages, skenographia employed highly sophisticated perspectival techniques which were designed subtly to modulate between reality and illusion in a variety of ways. For a largely illiterate Roman populace, theatrical performances provided a shared, mythological language which could be adapted to send and receive ideologically- and politically- coded messages in public under the guise of seemingly innocuous “festive entertainment.” Performers, popular audiences and the political elite became increasingly sophisticated at reading and manipulating this symbolic, lingua franca, and freely deployed its codes in other public fora, such as triumphs, funerals, and in circus games (see Beacham, 1999). Although this language increasingly permeated public discourse in these and other ways, we contend that, as the primary, popular, institutional locus of symbolic representation as such, the theatre remained the most important medium through which this shared, symbolic language was negotiated and refined in Roman culture of this period. Consequently, if we are to understand the complex aesthetic-ideological codes operative during Rome’s decisive transition from Republic to Empire, we must study how this language was aesthetically and performatively constructed in the theatre. The wall paintings can enable us to do so.

Through illusionistic confabulations of real and imagined spaces, skenographia played with the borders between reality and illusion, thus establishing the aesthetic—and thereby ideological—“frame” for this public language. Indeed, skenographia was directly and explicitly implicated in these aesthetico-political discourses by contemporaneous Roman commentators such as Vitruvius and Pliny the Elder. Theatrical skenographia is therefore a prime site for scholarly attention. But of the many temporary stages that, according to Vitruvius (5.5.7), were constructed each year at Rome, none survives; and the permanent stage facades (scaenarum frontes) that do remain contain no trace of their temporary skenographic elements.

Fortunately, in De Architectura 7.5.2, Vitruvius reports that Roman artists depicted these scaenarum frontes on the walls of Roman villas and townhouses. The eruption of 79AD preserved numerous such examples in the Vesuvian region, particularly in Pompeii and Herculaneum; a small handful of instances from elsewhere offer invaluable comparanda, e.g. the House of Augustus in Rome. These domestic adaptations of a public, theatrical art-form, then, represent an extraordinarily rich resource for the type of theatre-historical research conducted here.

However, prior studies of these paintings were hampered by the difficulty of determining and graphically representing different layers of depiction within the paintings. This was partly due to the illusionistic complexity of skenographic technique, but partly too, to the equally complex illusionism of the actual stage buildings they may be depicting: both temporary and permanent stage buildings included elaborate mixtures of real and painted architecture, and additional painted sets and embellishment (panels, trompe l’oeil effects etc.; cf. Vitruvius 7.5.5). It is thus readily apparent how difficult it is for the contemporary viewer to distinguish which elements of a given wall painting represent (i) painted illusionistic elements that may have been formed part of the decoration of ancient temporary stage sets, (ii) the actual physical architectonic structure of the stages themselves, or (iii) other illusionistic elements within the composition of the wall painting itself. In all this, of course, the interpreter’s difficulty is exacerbated by the fact that, unlike a Roman viewer, we have never seen the “real” scaenarum frontes from which, as Vitruvius notes, these paintings are often derived.

With the assistance of substantial grants from the Nuffield Foundation and the Getty Museum, therefore, Richard Beacham attempted physically to reconstruct two full-scale replicas of Roman temporary stages directly based on the evidence of specific wall paintings: “the Room of the Masks” from the House of Augustus in Rome, and Room 23 from the Villa of Oplontis, near Pompeii. These sets incorporated a system of movable painted scenery, also derived from the evidence of the wall paintings and ancient literary sources. In both cases, highly successful productions of ancient plays were staged upon the sets, empirically demonstrating for the first time that the wall paintings could indeed be used as evidence for workable ancient temporary stages.

This research produced many valuable insights and contributed to an evolving scholarly consensus that these paintings do indeed depict stage sets; a view previously hotly disputed (see e.g. Engemann, 1967, or Cerutti, and Richardson, 1989). However, this inordinately costly and labour-intensive process is not readily available to the numerous scholars in art history, cultural studies, classics, architecture and archaeology attempting to address the intriguing challenges presented by these provocative and unique paintings, and their environments. Meanwhile, the nature of the evidence for ancient skenographia and stage architecture provided by most of the paintings—which, in many cases are more complex illusionistic compositions—remained relatively un-explored.

These and other limitations can now be overcome by computer-aided, 3D visualisation. Both the ancient technique—skenographia—and the modern one of digital 3D modelling undertake the perspectival depiction of architectural structures: it was consequently a natural progression to conjoin them to extend our understanding of ancient stages. In doing so, not only did we learn much about the structures depicted in the paintings, but the detailed and analytic process required also revealed much that was not well-understood about the nature of the perspectival techniques employed by the ancient artists, and how these techniques influenced choices about the display (and viewing) of the works within their ancient architectural settings. Questions such as the relationship between painted artificial architecture and the actual architectonic qualities of the ambient domestic space; the manner in which sumptuous theatricalised décor may have been fashioned to evoke forms of public architecture to aid the Roman patron in the projection of his dignitas; or the probable structural and scenic qualities of the real stage buildings whose influence the paintings are believed to reflect and record, could not begin to be addressed until the qualities of the three-dimensional formats the paintings had perspectivally represented (and how they did so) was better understood.

In the first instance, therefore, the Project investigated the skenographic wall paintings by using digital 3D visualisation to create three-dimensional models of the stage structures they seem to depict.

There was a second, major component to the proposed Project. Studies had shown how, due to the hierarchical and deeply allusive manner in which domestic spaces and décor were configured in Roman culture, it is essential to view paintings in their original architectural locations in order to understand the range of meanings they may have produced. In many cases, however, this was impossible, since the paintings have been dispersed to museums around the world or, destroyed, are now known only through graphic records. Digital modelling can reunite painting and place. By digitally reconstructing the rooms where the paintings were originally displayed, and restoring to them their decorative schemes, we were able to analyse the relationships between these rooms and paintings within the architectural ensemble of particular houses, and discuss how they may have been perceived and experienced by ancient spectators. For example, research by Beacham and Denard (informed by such work as Bek, 1982, and Ling, 1999) suggested that a computer-aided approach would be extremely revealing if applied to the analysis of dining-rooms (triclinia) and adjacent spaces and their décor in a number of houses, in particular: the House of Siricus (VII.1.25 & 47); the House of the Centenario (IX.8.3); the House of the Triclinium (V.2.4); and the House of Marcus Lucretius (IX.3.5).

Research Hypotheses and Objectives

As artefacts of extraordinary historical importance, Roman wall paintings have received massive scholarly attention. But core questions remain relatively unexplored – particularly with regard to the paintings’ relationship to the “theatricalised” culture of the 1^st century B.C. In such paintings we see a conflation and merging of “real” architecture (painted of course) with the “artificial” architecture depicted upon panels contained within the stage-like structures represented by the artists. The spectator stands within the actual space of the room on the walls of which these paintings have been executed, and with whose architecture the structures depicted in the paintings have a direct connection, extending and linking up to the architectonic elements of the room.

Theatre painting was mimetic and playful. Vitruvius notes that skenographia had developed trompe l’oeil to create illusionistic elements which conflated the actual architecture of the stage façade with depictions of “real” architecture, and with paintings of “artificial” architecture displayed upon scenic panels within that architecture; sometimes the paintings’ fantastical compositions seemed to dissolve the reality of the actual façade altogether. Wall paintings sought to import this public, theatrical spectacle into the domestic sphere.

Scholars are now opening up exciting new avenues into the thematic and spatial “performativity” of Roman wall painting, and into the acts of spectatorship they attempted to precipitate. But, viewing domestic spaces as autonomous aesthetic ensembles, they underestimate the extent to which they may be evoking or adapting specifically theatrical configurations of space and spectatorship. The project researchers' studies of the relationship between Roman theatre and Roman decorative arts involves analysing how Roman theatre’s techniques of perspective, distance, reception, narrative, and the spectacular may be being replicated or appropriated in the domestic sphere, and led them to formulate the following hypotheses, subsequently tested through this research project:

The perspectival system (skenographia) used in such paintings can be “de-coded” to reveal and represent in 3D the architectonic structures depicted.
Through this process we may better distinguish between “fictive” and “real” elements depicted in the paintings, and interpret this to tell us about the actual stage structures and scenery that have in part inspired the paintings.
We can thereby attempt to identify formats of temporary stage buildings, and thus “see” in the paintings the actual theatrical referents until now “invisible” to contemporary eyes.
This will assist us in understanding how theatricalised conceptions of spectatorship and space have conditioned the creation of “performative” domestic paintings, their dispersal in the houses, and the behaviour of those viewing them.
Such work, focussed upon Roman theatrical practice, will advance and transform current debates about cultural signification, mnemonics, social hierarchies, and the conception and display of status, wealth and leisure in domestic Roman architecture and décor.

Following analysis of numerous skenographic paintings conducted by the project researchers, structures depicted in a selection of them were reconstituted in the form of three-dimensional computer models. When paintings suggested the presence of scenery—curtains in combination with painted panels—we sought to deduce plausible systems for the “rigging” and operation of such scenery through VR modelling. These models were compared in terms of their architectonic elements, proportions, and decorative details to attempt to identify generic characteristics of temporary stage sets. We compared these characteristics with the proportions and structural elements of facades of near-contemporaneous permanent theatre structures to determine what relationships may exist. Similarly, we investigated whether, and how, scenic elements visible in the paintings were used within the architectural format of permanent scaenarum frontes, or upon the scenic periaktoi placed before them. We also reconstituted and examined through 3D visualisation the positioning of a selection of the now-dispersed skenographic paintings within their original rooms. Using computer simulation, we analysed sightlines, architectural “vistas,” and perspectives created in some houses in relation to the use of such elements in the theatre, and considered how ancient viewers would have seen them as they moved through, or positioned themselves within the house.

Significance

The project is significant and novel at a theoretical level:

Broadly, through its contribution to current scholarly initiatives to move discussions of Roman wall-painting beyond centuries-old preoccupation with formalist analysis (classification of styles, determination of chronology, identification of presumed Hellenistic “old Master” models) to assess the paintings as part of a complex interplay of signs through which Roman aesthetic, social, cultural, and economic values were produced and negotiated.
Through its aim to reconcile two fundamentally different perspectival systems: ancient skenographia (not the same as post-Renaissance vanishing point perspective), and modern computer-aided 3D visualisation; each of which represents 3D architecture.
By its hypothesis that many of the structures depicted in Roman wall painting are based on actual theatre-architectural structures as referents, and that consequently, but speculatively:
These long-vanished actual structures may validly be reconstituted virtually using the wall paintings as evidence.
By similarly investigating whether paintings may be used to identify and reconstruct ancient movable scenery systems (scaenae ductiles/periaktoi); a significant scholarly “break-through” in understanding ancient theatrical technique.
Through its hypothesis that the location and organisation of the paintings within Roman houses create/reflect complex systems of cultural signification, mnemonics, social hierarchies etc., and that VR representation and evocation of the spatial dynamics of such architecture can help reveal such “performativity” for analysis.

The project is important at a practical level:

Through its intensely interdisciplinary nature: drawing upon and synthesising current scholarly work (much of it speculative/cutting edge) in art history, computer modelling, architecture, classics, cultural studies, archaeology, and theatre studies.
In developing a new methodology to analyse painted images and to convert these into the complex information categories and datasets required for CAD modelling of hypothetical “real” structures.
Through its use of new technology to “restore” lost or dispersed paintings to their original locations within particular houses reconstituted in virtual form.
Through its creation of 3D reconstructions in formats that can be observed by users in pre-rendered, high-detail animations or, alternatively, navigated and explored by them in real-time within lower-detail, but still architecturally accurate models.
Because it has produced new methodologies, enabling new ways of knowing, and indeed new types of knowledge which potentially are widely applicable in cognate areas of research in art history, cultural heritage, architecture etc.

Research Methods

The project researchers had already created a substantial database of source material including new photo-documentation (prepared during extensive work at Pompeii), historic photographs, ancient accounts, depictions of deteriorated or lost frescoes, and interpretations of Roman stage formats and scenic practices. With the experienced 3D Visualisation team in the School of Theatre Studies at the University of Warwick, (e.g. the AHRB-funded Pompey Project and Theatron EC), and drawing on the advice and collaboration of experts in Roman architecture, we created Virtual structures exploring the often complex architectural formats suggested by a selection of the paintings.

In this process, two-dimensional photographic images of ancient perspectival wall-paintings were transformed by a skilled modeller, closely collaborating with the Project researchers, into 3-dimensional digital format using architectural modelling software. The structural qualities of each architectural element were assessed to determine how the various components must have been used and integrated into edifices that could fit together and “stand up.” Scenic elements were similarly modelled and tested within the replica stage structure.

In selected examples where the painting had been removed from its original architectural location, we used digital, 3D models to recreate the original rooms’ decorative ensemble, and simulate other ambient qualities of the space including patterns of light and shadow which conditioned viewing, and vistas available to those in the space; standing, sitting, reclining, or moving. We investigated probable patterns of movement, and the hierarchical spatial organisation within the houses, and prepared interpretative materials to place them, their layout, and décor within appropriate cultural-historical and art-historical contexts situating our work in relation to contemporary scholarly discourses.

The Research Team

The research team consisted of: Prof. Beacham, Dr. Hugh Denard, Mr. Drew Baker, and Mr. Martin Blazeby. The 3D Visualisation Centre, beginning in 1996, has worked on a series of grant-funded projects, developing outstanding skills in computer modelling of historic theatre spaces, the collection, collation and multi-media presentation of associated datasets, and the development of sophisticated user-interfaces for accessing, manipulating, and interrogating the results of such research. Together with eight international partners in the EC sponsored THEATRON consortium (which Beacham planned and led: http://www.theatron.org) they completed an extensive inter-disciplinary project creating VR models of some 25 historic European theatres, together with elaborate supportive data and an interface for educational users. The Centre also worked on the Leverhulme and AHRB funded “Pompey Project”, a standard-setting programme of research in “virtual archaeology”.