A little over 150 years ago, a young British adventurer named Austen Henry Layard began excavations at two sites in northeastern Mesopotamia, in what is today Iraqi Kurdistan: at Kuyunjik, the citadel mound of the ancient city of Nineveh and on the citadel mound of Nimrud, the site of the ancient city of Kalhu (see the map and plan at the left; hover over to enlarge).

It was at Nimrud that Layard unearthed the best preserved of the ancient Palaces of Assyria, which we have come to call the Northwest Palace of King Ashurnasirpal II, an Assyrian monarch who ruled from 883-859 BCE (see a portrait of Layard at the left; hover over to enlarge). It is called the Northwest Palace because it was found near the northwest corner of Nimrud/Kalhu's citadel mound.

Layard worked for six years (1845-1851), excavating and documenting these sites. As at Nineveh, he and his assistants planned, drew and removed from the ruined site of the Northwest Palace the collapsed brick and stone bas-reliefs that decorated the palace's walls. There are a number of drawings that are testimony to the Layard years in Assyrian Mesopotamia, preserved in the British Museum and the British Library. Here on one of them he is shown sitting in the middle of his excavation documenting the bas-relief at Nineveh (hover over the image at the left to enlarge)....

and in another, at Nimrud, supervising the removal of two human-headed, winged bull figures--we have come to call them lamassu--from a gate in the central courtyard of Assurnasirpal's palace (hover over the image at the left to enlarge). Actually the bull on the left is in this museum. The drawing is inaccurate. I can explain why in the question period at the end of our presentation.

There are also sketches of workmen lowering bas-relief from the palace's walls (as seen at the left; hover over to enlarge).

Over the course of Layard's excavations and those of his Nestorian Christian assistant, Hormuzd Rassam, and with Ottoman approval and the approval, support and assistance of Henry Rawlinson, the British government's representative in the area, the stone bas-relief decoration was distributed to friends, family and the British Museum. Some of the relatives were supporters of his work in Mesopotamia and its publication, for example, his cousins the Guests. This is his cousin, Mrs. Charlotte Guest (hover over the image to enlarge). Most of the bas-reliefs in this museum, shown in the Assyrian World gallery on the second floor, are from the Guest collection. The story of how they got here can be read in a wonderful and to me aggravating saga written by John Russell, in his book From Nineveh to New York. What Russell tells is a good example of what happened to the visual depictions which Ashurnasirpal had created to decorate and give meaning to his life and to extend his personal fame forever. As he had his scribes write in his inscriptions, he built the palace "for the leisure life of his lordship."

Visitors came to Nimrud both during and after the concluding years of Layard's time in Mesopotamia and then in the years during which other excavators, including the British Museum's William K. Loftus and William Boutcher (1854-55) worked at Nimrud. These visitors received permission, literally, to mine the site of what were considered duplicate images, some of them fine examples of 9th Century BC bas-relief sculpture. This photograph is of the top of the citadel mound as it appeared during Rassam's second time at Nimrud, between 1878 and 1882 (hover over to enlarge). It is one of the earliest excavation photographs of Nimrud, and one of only a few surviving images of what the site must have looked like in the 19th century. You can see remains of pieces of the slabs of bas-relief or pavement on the right.

When American missionaries visited the sites of Assyria, which for them were both a curiosity and of religious interest, because both Nineveh and Kalhu are cities mentioned in the Bible, they came away with whole or parts of individual slabs. They sent them home to their respective colleges and seminaries from Virginia to Maine. This is documented in the records of the institutions that received whole slabs and fragments of slabs as gifts from these missionaries, often alumni of those American institutions. The records show that the bas-reliefs, if they were not already broken, were cut up into smaller pieces so that camels or donkeys could more easily transport them over land to where they were loaded on ships bound for Europe and then America. The letter on the screen is from the records of Williams College in Williamstown, Massachusetts, the first college to receive an example of a bas-relief in America (the letter is from D. W. Marsh to A. L. Perry at Williams College, 1882; image courtesy of the Williams College archives; hover over to enlarge). Layard, Rassam, and Rawlinson had also cut up the bas-reliefs for more easy transport away from the site.

Proof of this practice is everywhere evident at Nimrud in the extant remains of the Northwest Palace. This slab, from a room in the palace's southern wing, must have proven too great a task for removal, even though its upper half was broken when found, or got broken while it was being considered as an example worth taking (see the image of relief S-10 at the left; hover over to enlarge). In any case, the human head of the winged deity, the so-called genius, one of the protective divine figures that feature prominently in the iconography of the audience halls of the Northwest Palace, was chopped away from the rest of the bas-relief slab and found its way to the United States as a discrete piece of sculpture. As a piece of art is was treated as if unrelated to the rest of the figure. It is now in Chicago's Art Institute (see the image of the head at left; hover over to enlarge). In the old exhibition at the Art Institute, it was shown in a frame like a picture. Things have changed. Today it is shown without the frame and with an accompanying drawing to show its context. A complete eagle-headed version of this kind of protective figure, which stood to the right of this now-mutilated genius, is in this museum, in the Assyrian World Gallery.

The tracing of the movement--the buying and selling of these slabs over the last 150 years--has become a daunting task for those of us who are interested in the palace as a whole, or at least as Layard found it. Examples of bas-relief from the Northwest Palace are now to be found in 65 museums and private collections across the world. The largest collections of bas-reliefs from this palace outside of Iraq are in the British Museum (see the example at the left from the throne room; hover over to enlarge). Many of them are from the excavations of Layard, himself, and come from the palace's throne room. The descendants of Layard's circle gifted others to the Museum. In the 1950s and 60s when Max Mallowan and David Oates returned to excavate Nimrud on behalf of the British School of Archaeology in Iraq, Richard Barnett, the keeper of Western Asiatic Antiquities wisely added to the British Museum's collections by literally scooping up whatever he could find on the antiquities market. The second largest collection is here in the Metropolitan Museum of Art. That is the story in Russell's book, which I recommend to all of you.

Because there are several possible ways that bas-relief was taken from Nimrud, with or without permission, as has been the case all to often over the last 150 years, we do not know how many bas-relief fragments remain to be discovered in private collections. Every few years, another fragment surfaces. Today we can document the whereabouts of 321 complete and fragmentary examples of Northwest Palace sculpture. Tracking their movements can be a full-time job. When the British School returned to Nimrud after the Second World War, they re-excavated Layard's finds and extended their work into parts of the Palace that neither Layard nor Rassam touched, especially the private rooms beyond the south wing of the audience halls. They also began to restore bas-relief to the walls (see part of the British restoration efforts at the left; hover over to enlarge). The principal result of the British excavation work was the realization that the palace was much vaster than Layard's plans and reports indicated. When the British moved to work off the citadel in the lower town, the Iraqis continued to excavate, ultimately turning the palace into a site museum.

In 1974, the Polish Center of Mediterranean Archaeology, based at the University of Warsaw, arrived at Nimrud with a contract to excavate the area of the Central Palace of King Tiglath-pileser III which lies to the south of the Northwest Palace (see a view over the Polish excavation area at the left; hover over to enlarge). One of the by-products of the Polish time there (1974-76) was the attention that the director of the Polish project, Janusz Meuszynski, paid to the continuing presence of the Iraqi mission.

Meuszynski, with permission, had the whole palace site documented on 35 and 120mm film, photographing the restoration work (as in the restored lamassu at the left; hover over to enlarge), every fragment of fallen and broken bas-relief (as in the bas-relief example at the left; hover over to enlarge), and all the pavement slabs. These were added to another set of photographs that Meuszynski had taken in 1971. Meuszynski also arranged with the architect of the Polish excavation, Richard Sobolewski, to re-survey the site and record it in plan and in elevation. This became the first scientific plan and elevation since Layard's time.

To illustrate how much more we know about the palace plan 150 years after Layard's time, this is Layard's plan and and this is Sobolewski's (see the comparison at the left; hover over to enlarge). After the accidental death of Meuszynski in the spring of 1976, the Polish work at Nimrud ceased. Sobolewski became the caretaker of all of Meuszynski's records, finishing Meuszynski's publication of part of the palace and then turning to me [Sam Paley] to help with the rest. In 1992, we completed a restoration on paper, published by the German Archaeological Institute, of all the known evidence. Since then several new pieces of sculpture have turned up; they have either been published or are in the process of being published.

To explain how this paper restoration was done, each relief and fragment of relief, in situ and in the world's collections was drawn to scale. Identification of the position of each piece where it was thought to belong in the decorative scheme of the palace--was confirmed or researched. The work of many scholars--Cyril Gadd, Ernst Weidner, John Stearns and Julian Reade--over the last 50 years helped in the identification of the positions of each relief fragment. The re-assembling of the decorative motifs of room I in the east wing of the palace is a good example.

This is a portion of Richard Sobolewski's plan of Room I (at the left; hover over to enlarge) which I marked in colored pen. Complete, restored in situ bas-reliefs were marked in red, partially restored bas-reliefs in green and bases of the slabs in blue. This is how we started.

The motif in this room is in three registers (see an example at the left; hover over to enlarge): kneeling human-headed geniuses alternating with trees on the upper register, a summary historical and building inscription on the middle register, and eagle-headed geniuses alternating with trees on the lower register. There are some variations, but basically this is it.

An elevation drawing was then prepared, which showed what was there in situ and what was missing (see the drawing at the left; hover over to enlarge).

And then we figured out a hypothetical scheme, which took into account the direction in which each figure stood relative to the trees. We were helped by the fact there were enough broken figures on the in situ bas-reliefs to make a fairly good estimate of the total composition in the room (see the results at the left; hover over to enlarge).

And in several cases we were able actually to make some real joins. This is an in situ photograph of the slabs I-18 and 19 which joins a figure in the Lady Layard Collection in the Ashmolean Museum of Oxford (hover over the image to enlarge).

The next step was to assemble all the pieces that we could find belonging to this room according to the schematic, matching inscription fragments, iconographic details, sizes and shapes of the cut and broken stone (see the mockup at the left; hover over to enlarge)...

and then to paste up photographs to scale (see the mockup at the left; hover over to enlarge) which could be given to our artist, Helena Lewakova, who made the publication drawings.

The room can now be visualized as a whole piece (see the final drawing at the left; hover over to enlarge). This is what Layard described when the room and its partly fallen bas-reliefs were first uncovered.

Having now arrived at this stage of our thinking, it became important to find a way to make this work more accessible to our colleagues, students and the public. With access to the actual site not possible because of the present political conditions, with the far-flung distribution of fragments of the decoration of its walls, and with the danger to its existing preserved remains from natural environment, pollution and robbery, the new computer-based methods of publication and presentation which are increasingly available for educational purposes, the idea of building a virtual palace becomes possible and necessary if this great monument is to survive as part of our heritage. An alliance with Dr. Donald Sanders and his company Learning Sites Incorporated was forged at a meeting of the Archaeological Institute of America three years ago. We--and I mean Donald Sanders and Learning Sites, Alison Snyder, professor of Architecture at the University of Oregon, and Richard Sobolewski in Warsaw--began and have spent most of the last three years working on the model with an emphasis on the Great Northern Courtyard and Throne Room suite, which is the north wing of the palace.

Primary in our minds was also the fact that in over a century and a half since its excavation, there had been virtually no attempts to visualize the remains of the Northwest Palace of King Ashurnasirpal II of Assyria (883-859 BC) except through discussions, including our own, of the iconography and the historical and religious context of its bas-relief decoration. Most textbooks still reproduce Layard's commissioned drawing of 1849/50, in which he envisioned the throne room from the vantage point of somewhere near the middle of the room, that is, looking west toward Room C, an anteroom. That visualization has a distinctly Victorian aesthetic, which we have become used to: ceiling, light well, pastel shades and all (see Layard's commissioned rendering at the left; hover over to enlarge).

Our study of this interpretation of the remains in the context of other excavated palatial settings has led us to different conclusions. The new technologies we are using, including Virtual Reality, allow us much more latitude, both in testing our decisions quickly and in our ability to move around the "environments" almost as if we were actually at the site.

Our study has led us to different conclusions (compare the views at the left; hover over to enlarge). Among the problems we see in the Layard reconstruction are: (1) the major beams cantilevered from the walls to support the light well would most likely collapse, given that the opening seems to span the entire length of the Throne Room; (2) the light well, covering so much of the room, would seem to let in too much rain, sand, and blinding sunlight, damaging to the bas-reliefs, the people, and the mud brick; and (3) the people depicted are either too small or the bas-reliefs too big, at least compared to our current measurements (which also indicate that in the Layard drawing the room's proportions are incorrectly shown--if the width is correct, then the bas-reliefs are too large and the roof too low; if the bas-reliefs are correct, then the room's width is wrong).

The basis for all of this work is the recent architectural floor plan drawn by Richard Sobolewski, the architect of the Polish Center's 1970's excavations at Nimrud (see the plan at the left; hover over to enlarge). In his plan he took care to measure the precise sizes of each surviving architectural feature and bas-relief that still remained in place or could be assigned to a specific place along the palace walls because it was found collapsed on the floor directly in front of its position.

Using this plan, and a process that will be described to you in a few moments by Donald Sanders of Learning Sites, Inc., we worked to set up a proportional system based on Sobolewski's work, ancient historical precedents and modern architectural analysis to determine how high the walls might have been, the height and shape of the doorways, and the location and type of any decorative elements in brick or painted plaster that accompanied the carved bas-relief that adorned the lower walls.

Given the height differential between the interior and exterior bas-relief, we first surmised that the height of the throne room and perhaps the whole northern wing of the Northwest Palace (the principal throne room suite) was probably governed by the exterior façade of the throne room (see the throne room section at the left; hover over to enlarge).

This exterior façade is comprised of what Layard called courtyards D and E, and to which Sobolewski and I added and reconstructed a central doorway from extant remains found by the Iraqi excavations: the thresholds for the door, the seats for the door posts, and fragments of the lamassu door figures (see the plan at the left; hover over to enlarge).

This façade was really part of the southern wall of a great northern courtyard around which and in which much of the business of the palace that dealt with the outside world took place (see our reconstruction at the left; hover over to enlarge).

After examining the extant evidence of flat and arched doorways of various periods of Assyrian history (see the evidence at the left; hover over to enlarge) on the Nimrud citadel, at Fort Shalmaneser, Khorsabad, on bas-relief decoration and elsewhere, we adopted the shape of the arch drawn upon a painted and glazed brick panel fallen from above a doorway and discovered by the British at Fort Shalmaneser at Nimrud.

The doorway is dated to the period of Shalmaneser III, Ashurnasirpal's son and successor (see the panel and our proportional system at the left; hover over to enlarge). The brick panel allowed us to formulate a proportional system of bas-relief to arched opening. Then, from the increased size of the bas-relief flanking the central section of the throne room's outer façade, we surmised that the central doorway was tallest, with the parabola-shaped arch springing from the area of the Lamassu heads.

Once we set the height and shape of the central doorway at 8m and the side doorways at 6m, we set the height of the roof of the throne room façade so as to include room for the interior roof structure over the principal throne room suite. We surmised a parapet with typical canted crenellations at the top of the wall giving us a height of 12m for the façade from the paving surface to the top of the crenellations (see our drawn reconstruction at the left; hover over to enlarge).

Passing through the western archway of the façade from the courtyard, the archway was modified and became a (higher) flat ceiling on part of the interior passage, measuring 6.4m in height (see our section, looking west, at the left; hover over to enlarge).

Our reasoning for this was predicated on the remnants of rectilinear door elements found at Balawat (see the doorway from Balawat at the left; hover over to enlarge) and the hypothesis that there could have been a brick panel above the inside of the door such as the one at Fort Shalmaneser. Also, bringing the tall exterior arch to a flat arch allows for a height transition that works well with the smaller interior bas-relief panels. It also gives a nice rectangular pocket into which the doors close (see the section rendering above).

The proportions of the throne room, averaging approximately 10m wide by 48m long, seen in conjunction with our layout of the exterior courtyard façade, led us to settle on a throne room 10m high. Inclusive in the room are the wood beams (ca. 62cm square--actually in the drawing the beams are 62 X 51cm, since part of the height is buried in the ceiling) spanning this width and set every 3m along the length of the room.

To begin to suggest the grandeur of the throne room, a series of decorative motifs has been selected and placed along the tops of the bas-relief and over the doorways. They equal approximately the height of the bas-relief below. The idea for the layout of this decoration comes from the throne room at Fort Shalmaneser (see the examples at the left; hover over to enlarge). That room was repainted in the time of king Essarhaddon.

We also studied glazed and painted wall plaques, round and shield-shaped (such as the example at the left; hover over to enlarge), and

decorative elements from fragments of painted wall plaster and brick found by Layard and during the Iraqi and later British work in the palace at Nimrud (such as the examples at the left; hover over to enlarge), and

by Campbell Thompson in Assurnasirpal's palace at Nineveh (see some examples of the plaster decoration from Nineveh at the left; hover over to enlarge). At this stage of our work, we decided on a simple geometric decoration, which could be replicated easily, so that we could see what decoration would look like on the walls of our model. We are now considering how to include figural decoration for which there is a little evidence.

The idea for painted rafters with rosettes and circles with dots comes from the remains of the throne room in King Sargon's palace at Khorsabad (see examples at the left; hover over to enlarge). Therefore, it is thought that Ashurnasirpal's throne room B, the principal throne room, and Room C, the anteroom at its western end, would have been highly painted as opposed to a monochromatic scheme.

The throne room itself is not well preserved today and would disappoint both local and foreign visitors, the latter even if they could go there (see a view east toward the area that once held the throne dais at left; hover over to enlarge).

The actual bas-reliefs, most of which are in the British Museum (see two examples at the left; hover over to enlarge) and

the published drawings allow us to study them individually or as part of a decorative scheme. There are some bas-reliefs still in the throne room (such as the example at the left; hover over to enlarge).

Our model will have all the evidence and present a virtual environment that will be easily visited and studied. One will be able to study a relief against the background of its total environment in a virtual room (see an example from the Learning Sites 3D model of the throne room, illustrating how photographs are used as textures to obtain accurate replication of the wall narratives; hover over to enlarge).

We are still questioning aspects of the interior and exterior, including building structure, bas-relief attachment, decoration and color, natural light and manufactured lighting, drainage, brick sizes and interior and exterior surface materials. We are experimenting with 3-dimentional renderings of furniture (see our model of the king's throne and footstool at the left; hover over to enlarge) and

with the placement of figures in the environment (see the sample rendering of the Northwest Palace throne room at the left, prepared for display in the Metropolitan Museum of Art's Assyrian Gallery; hover over to enlarge).

Donald Sanders will now show how this works [ed. note: at this point in the presentation, Donald Sanders gave a live demonstration of the Learning Sites virtual reality model of the Northwest Palace, including real-time walkthroughs of the throne room and adjacent spaces, live internal links to the evidence we used for the model, as discussed by Paley, and discussed the 3D modeling process to the audience; those notes and the accompanying virtual world are no longer available. For more on Learning Sites continuing efforts to digitally reconstruct the Northwest Palace, see that project's homepage.].