Roman Chester - The Ghost Goes Digital

One afternoon, warm and bright, the Welsh Hills faint from the north west angle of the walls that encase the ancient city of Chester.  The singular realisation that what I had taken for granted as a favourite promenade around the city, was in part, the ghost of a great fortress; a Roman imperial legionary fortress.  Indeed, that most harsh and romantic of places, a fortress on the edge of empire.

Looking east from the base of the Water Tower,
towards Bonewaldesthorne's Tower,
the north west corner of the city walls.

That afternoon was in the late 1970s. I had developed an interest in 'the Romans' in childhood, and knew that I was living in a city that was frequently marketed to visitors as 'Roman Chester', though in truth there was little left to see.

At that time, Chester was nearing the end of a period that had seen its ancient historic heart torn out in the name of shopping and car parks (though more was to come).  A sore legacy that still haunts the city.  The famous Elliptical Building fell victim to the bulldozers.  Had it been a time of greater foresight and care, you would now be able to walk on its courtyard, ponder the substantial footprint of the walls that remained, and see the bases of columns that may have supported an elliptical balcony.  Nowhere else, in the remains of the Roman Empire, could you have seen such a thing.

Now, it exists only in a handful of photos, and the memories of field archaeologists who worked mere yards ahead of the bulldozers.

But aside from the shame of its loss, it begs the question of how much we are missing.  Even if I could visit the site today, it would not reveal the truth of the Severan insula, nor indeed the elevations for the Flavian, which in any case was never built.

Some 22 years beyond that summer afternoon, I would be contacted to produce computer models and illustrations for the archaeological report, based on the detailed and precise interpretations of Dr David Mason.  Even so, we found that what worked well enough as plans and elevations on paper, in 2D, did not always work quite so well in 3D.  Months of work were lavished on making as much sense of the recorded remains as possible, within the orders of Classical Architecture, and within the bounds of common sense.  The results precisely and clinically defined the form of the interpretation.

The Severan insula, from the south east

On being introduced to the local archaeological society, at a special lecture to accompany the publication of the report, there were gasps at the sight of it.  One archaeologist exclaimed that "Such a building would be more at home on the Palatine, not here."  All very cordial, but the reality was that after so much work, so much testing of forms, it would have been harder to get it wrong than to get it right.

The Severan insula - the east facade

There is no claim at all that what we did is exactly how it looked in reality, or in the case of the Flavian, is what was intended.  The past is as much a place of the imagination, as it is rooted in the evidence and all that can be extraploated from it.  In archaeological reconstruction, and especially in the case of using computer modelling and visual effects (what is often referred to as CGI, or erroneously as 'VR'), every scrap of evidence, every plan, every spot height, can be brought together in their correct spatial relationships to show how a site might have looked, and sometimes importantly, how it could not have looked.  That same piece of computer modelling can then be detailed and surfaced to resemble a photographic reality.

A powerful tool for interpretation and research, but minutes into my realisation that summer day of the late '70s, such digital magic was around 15 years into the future.  For that moment, I continued along to the castle and then the river, looking east and north in my excitement at the 'discovery', not least because of the sheer size of the fortress.  Too many questions, too many uncertainties, but I recall that the abstract label of 'Roman Chester', had transformed to a vast and spectacular fortress and hinterland, albeit lacking any real form, and I wanted to know what it might have looked like.

It's inevitable that those questions and uncertainties plagued the following years.  No amount of research could reveal the truth, because by then, most of it was gone.  The last people who could have told me what it looked like had been dead for over 1,500 years, and even they would have struggled to describe it in its early days, some 350 years earlier.

There was no single incarnation of Roman Chester; as now, the settlement evolved, changed, was built and rebuilt.  Presumptions of marching camps, vague hints of early campaign forts, the early timber and turf defences, to the great stone built fortress of the Mid Third Century; all are Roman Chester.  Which itself existed within the landscape of earlier settlements, and would continue to guide the shape of the city through to the High Middle Ages.

In 1993, after a decade illustrating astronomy books, and being recently the proud owner of Apple Computer's latest and greatest graphics workhorse (easily eclipsed now by an average smartphone), I wrote to the Grosvenor Museum, expressing my interest to bring Roman Chester back to life, but that I needed some help.

So began a series of discussions with the Keeper of Archaeology, Dan Robinson.  Discussions which ultimately included others, and which spanned the years to the present.  This introduction to our series of reconstructions of Roman Chester, cannot pass without acknowledging the insights of the archaeologists and historians I have worked and debated with: Dr Peter Carrington, Dan Garner, Dr David Mason, Dan Robinson, Tim Strickland, and Tony Wilmott of English Heritage.

It is their academic work and insight, and others before them, which forms the skeleton on which these reconstructions are based.

At a time that is so reliant on technological gimmick and bureaucracy, it is easy to overlook that supreme collaboration of the artist and the academic, collaborations that with current imaging and media technology, can reveal a startling and precise glimpse of our interpretations of the past.

In the 1950s, archaeological reconstructions by Alan Sorrell came to the attention of the Minister for Works, Lord Molson.  Schemes were being investigated to better present ancient monuments to the public, and on seeing Sorrell's work, Lord Molson saw the potential and believed such talent should be more widely employed by the Ministry.  In 1957, official approval was granted to such a scheme, and in an answer given in the House of Commons: "The Ministry of Works are anxious to enable the general public to visualise what ancient monuments looked like in the days when they were in use.  Mr Alan Sorrell has therefore been employed to carry out drawings ..."

A sentiment which is all the more valid today, given our expanding knowledge and the abilities of computer modelling; a melding of the art and science of archaeology with the art and science of CGI/visual effects.  There can be little that is more powerful than images and film which immerse the viewer in interpretations of the past, grounded in archaeology and historical study.

All very fine, except for one thing: the bits that are missing.  There lies the tripwire.  With a Roman military site, we have the luxury of the Romans' consistency for layout; forts and fortresses tended to follow a set pattern.  Even so, great caution is required.  In the case of Chester, we know where the praetorium (legionary commander's palace) would have been, and there were lengths of walls and an apsidal end hinting at a baths suite, but other than that, no clue at all of what this presumably impressive complex would have looked like.  Beyond the south gate, we know where the mansio was, but not what it looked like.  The tribune's houses are a matter for conjecture; we don't know what the bridge across the Dee looked like; the principia (headquarters building) is well-defined in plan but its detailed structure is also a matter for debate.  The angle towers around the defensive circuit are relatively straightforward to site (the south east angle is still there), but were they roofed or open?  There is no way to tell.

The more one stacks the negatives, the less case there is for even bothering!  Except that at a certain level, we can infer structures without defining their precise detail; we can place insulae in the context of the fortress, and the fortress in the context of the landscape.  We can apply any given lighting, accurate for the time and place, a particular sky and atmosphere, local detail, smoke, carts, columns of legionaries on the parade ground, and marching out through the arches of the east gate.  Clusters of civilian buildings, plots of land, trees dotting the river cliffs, and barges on the water.  A sense of place, and something that if not wholly accurate in detail, has something of the look of the place at that time.

The patchwork of the North Wall; at the centre of the photo is the Third Century section, complete with cornice

Let's take a more positive view: we know where the barracks were, and equally we know where they were not.  We have a reasonable idea what they looked like.  Equally, we have a good idea of the defensive circuit's appearance, the number of interval towers.  Some years ago, we debated at length about the height of the gate towers - two or three storeys, more perhaps.  Stand far enough away, and the difference between two and three storeys is noticeable but not startling.  Similarly, if we have the walls a metre or so too tall or short, it is not visually jarring.  Though luckily as we have a section of Third Century wall, complete with base courses and cornice, in situ, we don't have to worry too much about that.

Something as massive as the praetorium is a problem when we have so little of it left.  And yet at a distance, a modest structure slotted into the space is plausible enough.  Frustrating though it is, we have to accept that some parts of a site are lost forever.  If we choose to reconstruct that site (which we do), then care is needed to choose an angle and distance that respects the lack of evidence in those sections.

And so to putting it all together, all that has been discussed and mulled over for so long.  Arguably, there is no right time to settle down to complete a series of reconstructions of a site, especially one that is as complex and long-lived as Roman Chester.  As much as the sum of the archaeological work is a glimpse of the site scattered across those centuries, so a reconstruction now is a glimpse of how we perceive that site in the early 21st Century.  Inevitably, that will change; people may agree or disagree, but that's fine because reconstructions of such complexity are there to not only present ideas, but to open the debate further.  As one eminent archaeologist said at a conference in Chester in 2000, the outlines and broken lengths of buildings that are excavated, all once had full walls, windows, doors and roofs, and reconstructing them with 3D modelling is a way to not only test hypotheses, but to remind ourselves of the past reality of these ghosts.

Roman Chester - Mid-Third Century, viewed from the south east

The first three views, and indeed the ones to follow later, are about context; how the major sites relate to each other, and how they relate to our interpretation of the early topography of Chester.  In an overall sense, one can argue that were we to travel back in time and see the actual topography of Third Century Chester, the future landscape of our own time would be clearly apparent.

In this view above, looking along a south east/north west axis, we have a view which encompasses many of the key landscape features we know today.  In the foreground is the gorge between sandstone cliffs, characterised by modern Chester's riverside promenade, The Groves.  The amphitheatre is obvious, but where we know Grosvenor Park, the civilian locals back then would be living and working, in cluttered buildings and yards stretching east and north, and lining a road that our Foregate Street mimics, give or take.  The modern city crowds up to the city walls, which on the east and north are mostly the line of the Roman walls, but to the west and south are Medieval and later, and indeed beyond what we believe were the extents of the Roman west and south defences.  In the Third Century, those modern crowded areas would be clear and most likely scrubland, marked by an occasional path and patrol track.  Along the cliff line, cutback with quarries, are the narrow valleys of first Souters Lane (a little beyond the amphitheatre), and then Lower Bridge Street leading to the bridge.  The Roman bridge crossing was not exactly on the line of the Medieval Old Dee Bridge as we know it, but from this perspective, the difference would be slight.  What is thought to have been the parade ground, has for years been the Frodsham Street car park, at least in part, a level area at the foot of the walls where there is still to be seen a fine run of Roman masonry, along with some later Roman wall collapse.  Perhaps the most startling difference is that of Chester being an estuary or even seaside settlement.  Silting and reclamation in subsequent centuries have all but masked that, and yet the clues are there in the ancient river cliffs if one chooses to look and see.

Roman Chester - Mid-Third Century, the western extents

Although we mourn the loss of so much fine archaeology within the fortress itself, this view anticipates what we hope will be a bright future in furthering the study and interpretation of early Chester, for here lies some of the biggest mysteries that are yet to be properly addressed.  In recent years, what has for so long been referred to as the 'Quay Wall', was shown to be high and dry by historic tidal studies.  Even at periods of high tide, when shallow barges could conceivably be moored at its base, the problem of lifting cargo to the top, some 20ft and more, threw doubt onto the hypothesis.  Cranes perhaps, series of jettys and ramps maybe.  Or perhaps the Quay Wall was purely defensive.  Excavating what remains of the Quay Wall is surely one of Chester's great archaeological investigations of the future, and we'd like to believe, the not-too-distant future.

In this interpretation, we have taken the infilling of a road from the west gate to infer that there was no access through the Quay Wall for its entire length (though ultimately evidence may say otherwise), rendering it a purely and monumentally defensive structure.  That in turn practically demands a spur wall projecting from the north west angle of the fortress, to enclose the western extents; arguably, if there were no spur wall, the western extramural area would be open to the north, making the Quay Wall little more than an ornament.  With an eastern arm of the Quay Wall sweeping into and along the northern side of the defile, which is now marked only by a depression in Nun's Road adjacent to Nun's Fields, it is clear that there is at least circumstantial evidence for an enclosed western extramural area.  One of those things we have debated long and hard; lacking any further evidence, it's all we can do.  I took a draft version of this to the launch of Dr David Mason's 'Roman Chester: Fortress at the Edge of the World', published by The History Press late last year.  After the lecture, myself, Dr Mason and Dr Carrington looked at the image and briefly talked through this point.  Aside from the inevitable (and not for the first time!) 'your guess is as good as mine' approach to a structure, one thing is apparent: if this is correct, it suggests the port of Roman Chester, at least at this time, was not along here.  One might expect jettys along the course of the river to the south, around the bridge and upstream, but the main port may have been further north.  In fact given this scenario, it would have to be.  The dark foreground woodland is what we now know as Curzon Park, and the wide expanse of water between there and the far shore, is, in our day, The Roodee, Chester's racecourse.

Roman Chester - Mid-Third Century

The same view as earlier, but settling to a summer evening; barges are moored, smoke drifts from cooking fires, and a faint mist settles along the river. The parade ground is silent for another day, and the woods along the river soften into gloom.

There may be little left to see, at least little that is obvious, and we have lost so much, but it does not detract from the reality that was Roman Chester, in all its times.  It truly was the 'fortress on the edge of the world', and who knows how archaeologists 1,700 years from now will be viewing our Chester, let alone Roman Chester?

There is still much to do, and more reconstructions to finish and show, but for all the questions and uncertainties, this once great fortress and settlement, its port and hinterland, with all our perceptions of the buildings and the might of Imperial Rome, through good times and bad, was a place that folk like you and me, called home.