Realism and reality are pretty loaded terms within the wider discourse on archaeological visualisation (see Goodrick et al. 2004). Whether the piece in question is a painting, photograph, movie, standalone digital reconstruction or part of a larger interactive narrative, the use of labels such as authentic or realistic will spark hot debate regarding the quantification, legitimacy, authority, interpretation, extrapolation, and transparency of not only the outcome but also the data sources, author and the methodological and theoretical frameworks that informed its construction (Molyneaux 1997 & 1999). Despite the long history of discussion on archaeological visualisation these debates remain largely unresolved, in part because the term realism can be difficult to pin down in relation to archaeological themes, but also because the purpose of the outcome so often defines how these debates should be framed and presented within the visualisation (Ogleby 2005). This blog post does not aim to break any new ground on this subject, but rather through example, demonstrate a cross section of some of the key debates surrounding realism in archaeological visualisation utilising the 3D reconstruction of 'Clifford's Tower 1263' as an example.
'Clifford's Tower 1263' was created over two weeks to fulfil the requirements of one of the core modules on the Masters of Archaeological Information Systems at the University of York. The module required the creation of a visualisation on any subject, provided it had tangible archaeological or culture-historic significance. I selected Clifford's Tower as it is a prominent land-mark in an area dramatically altered by time. Throughout the process of creating the visualisation I grappled with a conflicting archaeological and text-based record, a limited personal skill set in digital reconstruction and balancing the requirements of academic rigor and research with public interfacing and entertainment.
Image 1: Overview screen-shot of 'Clifford's Tower 1263'
The outcome, as demonstrated above, took into account the larger settlement of the area, not just the tower itself, and afforded you the freedom to explore the tower, keep, and surrounding area up until the boundary of St Mary's precinct - the idea being to create context, and afford the viewer the ability to understand the interactions between key elements in the reconstructed landscape.
In retrospect the tight timeframe and my inexperience with 3D modelling and texturing meant the project could never be realised to its full potential, or be produced to the standard I had initially imagined. The final review of the project called into question the validity of many of my constructs - how does the viewer know which elements are based on archaeological fact, and which parts were creatively extrapolated? In the areas of conflicting data, how was this conflict shown? How does an appeal to realism impact the authenticity of the model? Does several layers of compounded conjecture, extrapolation and fabrication render this construction misleading or invalid? These concepts subsequently have been playing on my mind and as such I will seek to explore a number of them through this post.
Image 2 (left) and 3 (right): Demonstrating overview shots from the top of the tower and behind St Marys Precinct respectively.
Realism: The Good, The Bad, and The Downright Misleading
One of the key criticisms of 'Clifford's Tower 1263' related to its appeal to an unadulterated realistic form (even if it never actually attained acclaim for seeming realistic due to my inexperience in 3D modelling). This criticism has been discussed to death starting with inked site drawings and interpretive artworks, being reborn with the popularisation of site photography and now reaching a crescendo in the age of 3D modelling and virtual spaces (see Piggott 1978, Roussou 1999, Rua and Alvito 2011).
A cursory glance through both published articles and more casual blog posts reveals how complex and contested the idea of realism for archaeological visualisation is - on one hand it is often characterised as the pinnacle of illustrative work, the point at which the artist can pat themselves on the back with a smug smile of contentment upon hearing a consumer mutter in awe "that is incredible, I can't believe it is not real!". The flipside of this is the notion that when something becomes visually (or phenomenologically) realistic the consumer stops challenging its authority, and accepts the presented visualisation of the past as fact - a dangerous and slippery slope which potentially ends with the internalisation of misinformation and falsification. Further to this realism is often closely tied to concepts of authenticity and believability which further complicates the situation.
It is a challenge which is not so easily resolved, as there are as many arguments are there are counters, and for as many reasons there are to strive towards realism there are equal and opposing reasons for abandoning that ship in favour of more stylised waters. But before we can take a stance on this notion it makes sense to understand the scope and impact of realism and reality.
What's in a name?
The concept of reality or realism can be taken in a number of different veins, and as such the successful navigation through of their complexities often relies on pinning down exactly which aspect you are challenging. You could infer it to mean a photorealistic render of a setting in which the technical modelling, lighting and scene create a believable imitation of the real world. It could be interpreted as the visual presentation of real data - the visualisation being a realistic rendering of what is known whilst excluding or making evident the extrapolation. Or alternatively it could refer to the creation of a visualisation of any form which evokes the feeling, experience or memory of person, place or object - the creation of a real, tangible embodiment - a category which Clifford's Tower 1263 was originally aiming to capture. But these ideas are not only the domain of ongoing academic debate, but also have wider social implications for public engagement and understanding. These end-point consumers have developed their own frameworks and structures for navigating, understanding and generating meaning from archaeological visualisations - understandings which are often as diverse, conflicting and complex as those which formed the visualisation.
Realism: Is it really so real?
Realism as we will discuss in this section refers to both a philosophy and a school of aesthetics. To briefly summarise it can refer to an attempt to represent a subject matter truthfully, avoiding artificiality or implausibility, and as such covers not only the pursuit for a realistic aesthetic, but potentially also the pursuit of representing reality as the subject. This definition has some obvious tensions at face value and is deeply problematic in reconstruction, interpretation and visualisation for archaeology as so often we only have fragmentary remains which necessitates the extrapolation of forms, ideas and meaning (Isenberg 2013; Rahaman, Das and Zahir 2012).
The aesthetic side of the discussion, at face value, appears to be the easiest to resolve. The closer to reality a visualisation appears, or feels, the greater its quotient of realism. In other words a high resolution photograph of the real world would represent a high fidelity of realism whilst abstract expressionism would most likely represent low fidelity realism. This division becomes a touch muddied if you allow a wider definition of realism or a discourse of its constituents. Or if you enter into a discourse about the reality of what is being represented realistically. In the case of the latter we could identify a photograph as being realistic, but also identify that the realistic rendering is of a staged, framed and static moment and as such is not necessarily realistic, even if it is aesthetically photo-real. Regarding the former we run into issues if we accept realism can supersede the purely visual and take into account concepts of persistence or presence (the feeling of a virtual world being real, and the feeling of actually being there). The discourse on presence and persistence is something I am hugely passionate on being an avid Oculus Rift user, but it is an issue which extends far beyond the scope of this paper and as such we will move on.
The philosophy of reality and realism posits some difficult quandaries by proposing that differential realities or interpretations of reality may be able to co-exist in any number of given forms. In other words, we may internalise and project reality differentially, and as such any given visualisation will be a representation of the base media which is then viewed and constructed into a visualisation by the author which is then understood through the reality framework of the consumer (Daston and Galison 1992). As such a shared, empirical standard for realism may not be attainable in the present, let alone when attempting to visualise the past. In light of this it may well be tempting to fall into existential nihilism, however hold onto some optimism for a bit, as it may also be possible to occupy a tenuous middle ground in which we understand visualisations as a reflection of compounded realities. The outcome of this discussion is that reality and realism may be better understood as constructs, and that rather than pursuing an empirical standard for the portrayal of reality in visualisation it might be more worthwhile to instead seek to understand the frameworks and structures which informed the various lenses of construction.
In the case of 'Clifford's Tower 1263' there was a concerted effort to create presence alongside an attempted aesthetic realism by avoiding stylisation, implausibility or the supernatural. However, the representation has a significant amount of creative construction, meaning it is not necessarily a truthful visualisation as it contains significant amounts of artificiality. However, if we were remove all the elements of conjecture we are left with almost an entirely blank slate. Interestingly, whilst this overcomes the issue of artificial construction, it creates a whole new one: artificial reduction. Which leaves us stuck between a rock and a hard place: on one hand we avoid artificiality by removing anything fabricated, but in doing so we cross over to the other hand where we are forced to interact with an artificially reduced image of the past. There is no hard and fast solution to this as a correct judgement rests on the purpose and parameter of a given work as well as in the way which we frame and implement ideas about realism and reality. In the case of 'Clifford's Tower 1263' the goal was to create a realistic feeling of being in the precinct, a brief which leant heavily on the processed reality approach discussed above and somewhat ironically required significant fabrication to produce an outcome which felt and looked more real. If the brief had been to represent a truthful reality based on archaeological remains alone a more appropriate outcome would have been to appeal to an artificially reduced reality.
Image 4: Demonstrating what Clifford's Tower Bailey would look like when all structures and assets not directly evidenced are removed.
To summarise: the notion of realism in archaeological visualisation is difficult as it draws not only on pure aesthetics but also on concepts of authenticity, truthfulness and artificiality. If we adopt a position of absolutes - taking realism to mean only aesthetic, or only accuracy, the tension can be nullified on one side, but at the cost of exponential issue compounding on the other. There is also the potential to take a liberal stance on philosophic aesthetics and realism - by accepting that representations of reality will be reflections of the author and the physical we can circumvent issues of absolutes, a solution which whilst workable, still requires discussion, and may not sit well with those of quantifiable dispositions.
Representing Real Data
As mentioned in the introduction, one of the key criticisms of 'Clifford's Tower 1263' was that there was no way to tell between what was well evidenced by the archaeological record, and what was extrapolated. In other words what was 'real', what was 'probably real' and what was 'probably not real but it looked good'. Using two innocuous houses from the model we will explore some of the ways to deal with this quandary, however, as we have become accustomed to over the preceding paragraphs there will not be any hard or fast resolutions.
Image 5 (left) and Image 6 (right): Demonstrating before and after asset removal.
Image 5 demonstrates the scene setup - detritus from every day work and life scattered around the houses - boots at the door, cunning chickens taking advantage of dumped grain sacks, tools for tomorrow's work waiting for the new day. The scene tells a story. A story which whilst based on evidence from comparable settlements has no record in the Clifford's Tower area. So in Image 6 they have been removed - leaving a sad, empty scene, albeit a better evidenced one.
The textures themselves pose somewhat of an issue: the brick textures from the walls were originally taken from a photograph I took in the Tower precinct. But these are bricks which have endured 800 years of weathering, replacement and human modification, so not the most realistic representation of how they may have looked in 1263. As such I made a trip to the mason's market at the Minster to take photos of fresh hewn stones only to discover that the stone was cut in different sizes, from a slightly different stone type to what I required. The resulting textures are a combination of the photographs taken of the tower, the masons market and excessive amounts of Photoshop. So if we are going for realistic representations of fact we should probably do away with those to leave an image un-marred by artistic liberties in texturing. As image 7 demonstrates this solves one issue of realism at the expense of another - the problematic, and unrealistic textures have been done away with, but we are left with an equally unrealistic grey-scale world.
Image 7: Demonstrating the removal of ambiguous textures
Removing the textures only goes so far to providing an understanding of which elements are well evidenced, which have some evidence and which are artificial extrapolations. To try to overcome this issue a transparency was applied to anything which was not well evidenced, and colour codes applied - green representing well evidenced material, transparent orange for those where there is some evidence but not enough to be sure of total form, and red representing areas that are extrapolated.
Image 8: Demonstrating the codification of ambiguity
Whilst the codification presents an interesting assessment of the liberties I took in the visualisation it lacks the experiential and immersive quality of the original visualisation by substituting the realism of archaeological fact for a more realistic expression of a possible past. I suspect that the decision about the success of any of these outcomes would be determined by the purpose of the visualisation. If the purpose of the visualisation is to understand the variation of certainty in the reconstruction then Image 8 would be a roaring success, however if the purpose was to create an embodied experience of a possible past reality then Image 1 is probably more appropriate. Whilst there may not be the ability for these to co-exist in one image, there is the possibility for a series of visualisations to provide a more comprehensive intersection if the demands of the project afford it.
Realism: A Story About How I Both Succeeded and Failed In Justifying It
The decision to produce a piece in a realistic fashion was informed in no small part by the desire to try provide a phenomenological experience of the site which is impossible to gain in the physical world now due to modern interventions - and whilst it is certainly possible to feel presence without realism (ie: a piece of surrealist artwork or a low-poly video single-colour video game has the same power to provide an experience of place, time or event) it was decided that a realistic image was required to provide the immersed in time effect. The first question that I was posed was: how would this realism impact on public perception and the potential propagation of misinformation as fact?
To try to circumvent the issue of public consumption of misinformation in 'Clifford's Tower 1263' I implemented a start-screen which informed the viewer they were about to partake in a world which, whilst informed by archaeological data had several layers of conjecture, and as such, whilst it aimed to be believable and based on facts, it was probably not a very realistic depiction of the area. I thought this was a pretty elegant way to hold the hand of the viewer and make sure they walked away with a better understanding of Clifford's Tower and its context, but also knowing that what they had experienced was an artistic interpretation.
This solution seemed to appease the great archaeological overlords who helped me develop the theoretical backdrop for the work, but the reaction of family and friends (who are not archaeologists) was quite to the contrary. Their reactions spanned from "so you mean this isn't real? Why did you bother making it? And why am I bothering to play it?" to "of course it isn't real... " through to "I don't care, I just want to see what it looked like". Whilst comical in retrospect these kind of responses highlight how a seemingly innocuous decision to inform patrons had ongoing repercussions in some instances, and was a complete non-issue in others.
In short, issues of realism, or making evident the facade is a deep seated issue in public presentation which is linked to ideas about the role of archaeologists, the legitimacy of their interpretations and the reliance on realism as a mechanism for making images relevant. For some, no matter how realistic an image looks they will always frame it as an interpretation. But for others saying the reconstruction's reality was a facade that called into question my legitimacy as an archaeologist, as well as the relevance of the reconstruction to understanding the past.
There is no neat conclusion to this section, only more questions. Realism in archaeological visualisation is a tricky subject that has a wider context and influence than the isolated framework of academia - navigating these issues, as with the rest of the questions posited in this blog post, most likely will never have a static end point, but rather need to be considered on a case by case basis with reference to the current state and the desired directions of a given visualisation.
This blog post overviewed just some of the issues inherent in the concept of realism and reality for archaeological visualisation. Although no solid conclusions can be offered to resolve the complexities and conflicts it has been shown that by making explicit the purpose and paradigm of a visualisation a tenuous middle ground for academia and public engagement can be explored. In the case of 'Clifford's Tower 1263' the form of the outcome was dictated by the function - and as such the criticisms regarding realism and reality, whilst valid, can be explained and justified. The topic of public engagement and the way in which reality and realism are framed by the consumer are incredibly complex strands of the debate which require greater attention that I can afford them here. Looking towards the future it would seem that whilst we may not be able to reach an end point and agree on the best way to handle realism and reality in visualisations, we can at least mitigate a number of the prevailing criticisms by making explicit the function, and addressing how the form fits with this.
Daston, L. and Galison, P. (1992.) The image of objectivity, Representations 40, 81-128.
Goodrick, G., Earl, E., Huggett, J. and Ross, S. (eds.) (2004). A manufactured past: virtual reality in archaeology. Internet Archaeology, 15.
Isenberg, T. (2013). Evaluating and validating non-photorealistic and illustrative rendering. Image and Video-based Artistic Stylisation, 311-331.
Molyneaux, B. L. (ed.) (1997). The Cultural Life of Images: Visual Representation in Archaeology. London: Routledge.
Molyneaux, B. L. (1999). Fighting with Pictures: The Archaeology of Reconstructions. Journal of Mediterranean Archaeology 12 (1), 134-6.
Ogleby, C. L. (2005). The 'Truthlikeliness' of Virtual Reality Reconstructions of Architectural Heritage: Concepts of Metadata. In The International Archives of Photogrammetry, Remote Sensing and Spatial Info Information Sciences 34, 192-9.
Piggott, S. (1978). Antiquity Depicted: Aspects of Archaeological Illustration. London: Thames and Hudson.
Rahaman, H., Das, R., & Zahir, S. (2012). Virtual heritage: exploring photorealism. Progress in Cultural Heritage Preservation, 191-200.
Roussou, M. (1999). Incorporating Immersive Projection-based Virtual Reality in Public Spaces. Proceedings of the 3rd International Immersive Projection Technology Workshop, 33-39.
Rua, H. and Alvito, P. (2011). Living the Past: 3D Models, Virtual Reality and Game Engines as Tools for Supporting Archaeology and the Reconstruction of Cultural Heritage - The Case Study of the Roman Villa of Casal de Freiria. Journal of Archaeological Science 38, 3296-3308.