The Visual Model, from Above


Our goals for creating the visual models of St Paul’s Cathedral and the buildings surrounding it in Paul’s Churchyard have been to achieve the highest possible degree of accuracy and authenticity in our work. Nonetheless, due to gaps in the evidence, we have had, from time to time, to infer various aspects of our models from the data we have or — when only minimal data is available — to use models representative of the genre or type of building we know or believe to have been in that spot.

One way of thinking about the visual model is to imagine it as an exercise in data visualization, not unlike a pie chart or a bar graph. In this case, however, the number and kinds of sources, their relative accuracy and authenticity, and their incorporation of nonrepresentational conventions of display exceed significantly the number of variables usually integrated in conventional data visualizations.

A good bit of information survives from pre-Great Fire London to help us achieve that goal. Unfortunately, the data we have is often partial, fragmentary, contradictory, or imcomplete. On a number of occasions, we have had to fill in missing portions of the data with what we hope to be informed guesses.

A good example of this is our model of the Paul’s Cross preaching station, visible on the website of the Virtual Paul’s Cross Project. Thanks to the work of Francis Penrose, Surveyer to the Fabric of St Paul’s in the late 19th century, we know with high accuracy the size of the base of the structure.

Penrose excavated the foundations of Paul’s Cross and published his findings. To read Penrose’s full account of his survey of the Paul’s Cross foundations, go here: Penrose Archaeologia (1883)

Penrose’s measurements only give us a design for the base of the structure and a sense of its overall scale. To move upward from the base, we turn to images of Paul’s Cross that survive survive from the early modern period. Although they differ in many details, they give us a pretty good idea of what it looked like.

What we don’t know, however, is the height of the building. The height we finally decided on is based on an assessment of the overall structure with regard for proportion, overall scale, and a sense of what height was needed for the structure to function comfortably as a setting for a two-hour-long event.

Paul’s Cross, the Preaching Station. Model by Josh Stevens.

As this discussion suggests, the data we have available, while considerable, does not provide everything we need to know to produce a completely data-driven model. Although everything one sees in our model is based on a careful consideration of the historic record, there remain gaps to close, conflicting evidence to evaluate, choices to make, assumptions to ponder and question.

The models of buildings one sees on this website embody data that exhibit a range of levels and types of accuracy, of specificity or general categorization, of levels and types of accuracy.

We can sort the kinds of data available to us into the following categories:

1. Data-based accuracy — information that is accurate in the sense that it is based on actual measurements, like measurements of the cathedral’s foundations.

2. Historic accuracy — information that comes to us from the historic visual record, which is only as good as the image is accurate. After all, Hollar’s drawings do not always agree with his engravings. Which should we follow, if we have both?

3. Representational accuracy — when, as is the case for many of the buildings in the Churchyard, we have the measurements of the foundations of the building, and maybe a detail or two about the number of stories or number of garrets, but no further details. But we know from other evidence what kind of building it was, and we know what features buildings of that kind had, so we can create a representation of that kind of building based on other buildings of that kind for which we have images or can look at surviving examples.

4. Informed guesswork — what we do when we know some things, but not others, so we have to make a guess about the missing details.

Data Still to be Examined

Building our model has always been a work-in-progress. As we worked through the data available to us, we also from time to time became aware of sources of data with which we had previously been unaware. Some of this new information has been incorporated into the model. Other sources, however, came to our attention too late to be included. We also believe there is material yet to be found, all of which will shed further light on the structures inside Paul’s Churchyard.

Please see the material under the tab “Further Work” for a discussion of sources of information of which we are aware but were unable to take advantage of, due to limits on our available time, staff, and money.

Modeling the Cathedral

We started our modeling literally from the ground up. The Cathedral’s original foundations survive in the ground atop Ludgate Hill. They have been surveyed by archaeologists who have determined with a very high degree of accuracy the basic shapes and dimensions of the structure’s foundations.

John Schofield, St Paul’s archaeologist, has gathered all this data into his monumental study St Paul’s Cathedral Before Wren (English Heritage, 2011), a volume absolutely essential to our work.

Paul’s Churchyard, c. 1500. Fom Schofield, St Paul’s Cathedral Before Wren (2011)

In addition to the basic floorplan for the Cathedral and the basic geometry of the Churchyard itself, we have drawn on an unusual source for our measurements of the heights, Christopher Wren himself,who produced drawings showing sections of the Cathedral’s interior and a partial plan of the Cathedral in 1665 as part of his plan for remodeling the building. Wren thus provided us with the measurements for the shape and height of the arches, the windows, and the ceiling of the building’s various sections.

Wren was in the right time and the right place to be hired to plan the rebuilding of the Cathedral after it was destroyed by the Great Fire of London in 1666. One can see in this drawing that the style and some of the features of Wren’s St Paul’s were already in his mind, regardless of what one thinks about the integrity of plopping a classical domed structure on top of a medieval building.

Christopher Wren, Scale Drawing of pre-Fire St Paul’s Showing Proposed Dome (1665). Image courtesy Warden and Fellows, All Souls College, Oxford

After the details we learned from the work of archaeologists and from Wren’s drawing, the most important source of information about the look of pre-Fire St Paul’s has come from the remarkable set of engravings that the Bohemian artist Wenceslaus Hollar prepared in the 1650’s for the English author William Dugdale, as illustrations for Dugdale’s History of St Paul’s Cathedral (1658).

Hollar’s set of engravings covers both the interior and the exterior of the Cathedral. They have been invaluable to us in imagining details of the Cathedral’s design and appearance.

A small number of Hollar’s original drawings survive as well, most notably this image of the Cathedral’s east front and the following image of the Cathedral’s south side.

This image of the Cathedral’s south side, however, when compared to the engraving Hollar prepared for Dugdale’s book, points to the challenge sometimes presented by the surviving visual record.

While Hollar’s drawing is superficially identical to his engraving, we found that upon closer examination, differences between the two begin to appear.

For example, did the Cathedral — in the design of the facade of the South Transept — have one (the drawing, on the left, below) or 3 (the engraving, on the right, below) rounded-top structures over the main door? Were the panels to the left and right of the main door narrow without decoration (the drawing) or wider and decorated with carved panels (the engraving)? Were these panels narrow (the drawing) or wide (the engraving)?

This specific instance of variation between the drawing and the engraving — both by the same artist — was not important to us, because this part of the Cathedral shows, in Hollar’s drawing and engraving, a facade that is part of Inigo Jones’ neoclassical remodeling of the Cathedral from the late 1630’s, thus not a feature of the Cathedral as Donne would have known it. But the point here is that the visual record, the collection of images of St Paul’s Cathedral that come down to us from the early modern period, is not to be taken on face value, without careful consideration.

John Gipkyn’s famous painting of about 1616 showing Paul’s Churchyard during the delivery of a sermon at the Paul’s Cross Preaching Station, for example, shows a radically truncated Nave to the painting’s right, as well as missing bays in its depiction of the Choir, not to mention its giving the Choir Norman windows instead of the Gothic pointed arches that Hollar shows in his images of the Choir.

The Hollar images that show the exterior of the Nave and West Front also present the challenge of showing us not the facade of the Norman nave Donne would have known but the appearance of the building after Inigo Jones remodeled it in the mid-1630’s in a vaguely neo-classical style.

Wenseslaus Hollar, St Paul’s Cathedral, The West Front

This is especially noticeable in his design for the West Front, but it is also the case with the North and South facades of the Nave. Hollar’s image, below, shows the Cathedral’s Chapter House in the center of the image, but shows Jones’ remodeled Nave facade to the left of the Chapter House and what we presume to be the Cathedral’s original Norman facade to the right of the Chapter House.

We, of course, based our configuration of the north and south facades of the Nave on Hollar’s treatment of this part of the South Facade.

Our goal in all areas of the Cathedral in which Jones covered the historic facade has been to see what design lay beneath Jones’ work. But, as we will discuss when we get to the Cathedral’s West Front, on some occasions Jones’ work is an important source of clues to what we believe to be the original design.

Other early modern images of St Paul’s Cathedral and its environs help us fill in details. The image below, for example, from the “Copperplate Map,” shows the cathedral around the middle of the 16th century, with its spire intact.

St Paul’s Cathedral, Detail from the Copperplate Map, c. 1560

The image below also shows the spire before it was struck by lightning in 1561. This image is important because it is one of the few to give us a glimpse of the West Front before Inigo Jones got his hands on it in the 1630’s.

Prospect of ye Citye of London, c. 1575

The image below, from the Agas Map of 1563 shows the Cathedral after the ligntning strike and fire that destroyed the spire.

The image below is from a panoramic view of London by Wenseslaus Hollar, dating from 1647, which reminds us that medieval St Paul’s dominated the skyline of London in the early modern period even as Wren’s cathedral does today.

Data from Other Sources

We have also sought to incorporate information not usually included in visual models of historic sites. This includes, especially, information about the relative ages of the buildings and the times of day and the kinds of weather one would expect to find on the specific occasions we are recreating.

pc no crowdl
Paul’s Churchyard, the Cross Yard. From the Visual Model, constructed by Joshua Stephens

For illustration, the image above, from the Virtual Paul’s Cross Project, shows our model of the cathedral’s Cross Yard, with the Paul’s Cross preaching station in the center, the Sermon House against the cathedral to the right, and the houses surrounding the Cross Yard to the left. The image below shows the same view after the addition of details that reflect the relative ages of these structures and the look of a chilly, overcast day appropriate for outdoors on a chilly day in London in early November.

courtyard render 2 black roof WWW
Paul’s Churchyard, the Cross Yard. From the Visual Model, constructed by Joshua Stephens, rendered by Jordan Gray. 

The look of the houses in this image brings us to another topic, our understanding of how buildings of this type would have looked in the early 1600’s.

We are familiar with this way of imagining the external appearance of early modern houses, with their supporting timbers exposed, because we are used to it from our experience with surviving structures of this type in England. For instance, the image below, which shows the facade of the Black Swan Public house in Yorkshire.

The 15th century half-timbered Black Swan Public house, Peasholme Green, York, Yorkshire, England, United Kingdom, Europe

We followed, for the Virtual Paul’s Cross Project, the current practice of imagining early modern residential and commercial structures as having exposed beams painted black or brown, in contrast to the plaster facades, often painted white or in pastel colors, that fill in the space between the timbers.

Houses in the Virtual Cathedral Project do not, however, look this way. After reviewing visual images of similar structures from the early modern period, for example the ones shown below, we found that the overwhelming majority of these images show buildings of this type as having plaster over the entire exterior of these buildings.

As a result, we have come to the conculsion that in the early modern period these buildings were plastered all over, not just between the structural timbers.

For discussion of this issue, see specifically Waldemar Komorowski, “Layering of Facades: A few comments on the colour of Krakow’s facades in earleir and contemporary times,” as well as the much broader discussion in John Schofield’s Medieval London Houses (Yale, 1995).

What this means, of course, is that the exposed-framing style of construction is a modern reinterpretation of early modern construction style, a reconstruction that ought to have a post-early modern origin-point and a history all its own. We look forward to reading such an account.

Representational Modeling: The Case of the Bishop’s Chapel

We also know that the Bishop of London had a Chapel that sat against the north side of the Cathedral, near the west front, adjacent to the Bishop’s Palace. We also know that, by the early 16th century, there was a large Cloister situated against the north side of the Nave, between the Bishop’s Palace and the North Transept, and south of the Hall of Minor Canons.

We also know that in the latter part of the 16th century this Cloister was demolished. We also know that after this Cloister was demolished, structures were built against the north side of the Nave. These structures served as workshops for tradesmen and craftsmen.

We know the size of the Bishop’s Chapel and the approximate era of its construction, but have no images of its appearance. We do not know the size or precise location of the workshops.

So the image above is representative much of our work, combining precise data (the location and size of the foundation of the Bishop’s Chapel), representational data (the appearance of a chapel of this size and era of construction), and guesswork (the location, size, and appearance of the workshops).

St Paul’s Cathedral, the West Front

The most challenging part of modeling St Paul’s Cathedral in the early 17th century has been reconstructing the Cathedral’s West Front. This is because we were unable to find a complete visual depiction of the West Front as it looked before Inigo Jones’ remodeling work in the late 1630’s.

As a result, our visualization of the West Front constitutes the most speculative aspect of our visual model. Nonetheless, it is still based on a caareful review of the information available to us.

We do have partial views of the West Front which provide us with some clues as to the West Front’s appearance.

Prospect of ye Citye of London, Detail (c. 1575)

The image above shows the distinctive array of three tall windows in the facade, with the center of the three taller than the ones to its left and right.

St Paul’s Cathedral, Detail (West Front, from the northwest)

This image shows the circular opening in the triforium and (perhaps) the top of the center of the three windows shown in the Prospect image.

Civitas Londinium, detail (c. 1550). Woodcut, Pepys Library, Magdalene College, Cambridge

Finally, this image at least suggests the circular window in the pediment and the long central window in the facade, below the pediment.

These images all show the West Front to contain a central rectangular form topped by the pediment and flanked by towers on each side. These images all display a configuration of multiple vertical windows in the upper section of the West Front, topped with a triangular pediment.

Two of the three images show a round widow located in the triforium.

St Paul’s Floor Plan, the West Front (from the Codrington Plan)

We also had visual evidence of the configuration of the West Front from this drawing of the Cathedral’s floor plan, showing the size of doorways, plllars, and other architectural features.

We then turned to assessment of the very problematic evidence provided by the visual record of the later seventeenth century.

Wenseslaus Hollar created a very detailed image of the West Front for Dugdale’s History of St Paul’s Cathedral, but it shows the West Front as recreated by Inigo Jones in the 1630’s, not the facade of the Norman cathedral.

Wenseslaus Hollar, St Paul’s Cathedral, The West Front

Jones himself left us a preliminary draft of his plans for the West Front.

Inigo Jones, St Paul’s Cathedral, West Front

In spite of the imposition of Jones’s design, these images of the West Front show the persistence of the basic configuration of a pediment with a round window in it, over a rectangular facade, flanked on each side by a tower.

They also show a row of three windows, in which the center window is taller than the ones that flank it to the left and right. Both images of Jones’ work also show a large central doorway flanked on each side by smaller doorways.

We concluded that these elements would need to be included in our revisioning of the West Front.

We then turned to Richard Halsey’s comment on the design of the interior of St Paul’s nave (in his essay in St Paul’s Cathedral Before Wren on “Placing St Paul’s in the development of English greater churches” (pp. 233 – 236), that the nave “appears to be a design of the early 12th century that ultimately owes its three-storey elevation to Saint-Etienne, Caen.”

Fortunately for our purposes, this building survives so one can compare its interior design with Hollar’s image of St Paul’s Nave.

Saint-Etienne, Caen, the Nave
St Paul’s Cathedral, the Nave. Engraving by Wenseslaus Hollar, 1658

With Halsey’s insight about the interior of medieval St Paul’s as our guide, we discovered that the exterior of Saint-Etienne, Caen, looks like this:

St Etienne, Caen, France, the West Front

Using this design as our guide, and incorporating the design elements for which we have strong evidence were part of St Paul’s West Front, we came up with this design for our model.

St Paul’s Cathedral, the West Front

The result is a design that is more austere than other English cathedrals of this period. It does, however, incorporate the essential features documented by the 16th century visual images, together with the general style of St Etienne, in Caen.

Our sense of the austerity of this design is influenced by our knowledge of Norman cathedrals in England which have more ornate or complex designs. Part of that ornateness comes from the fact that many of them have been remodeled since their initial completion.

Norwich Cathedral, West Front
Winchester Cathedral, West Front

Perhaps the best example of a surviving Norman cathedral with which to compare our design for St Paul’s West Front is the cathedral at Rochester.

Rochester Cathedral, Engraving, c. 1650

At Rochester, plans for a remodeling of the nave were shelved, preserving the original, austere Norman West Front. While slightly more ornate than our design for St Paul’s, this example reassures us that if we were actually able to visit London in 1600, we would recognize the West Front of St Paul’s.