Göbekli Tepe, Turkey

Preliminary Site Conservation Inspection and

First Mortar and Plaster Documentation Report

12-24 September 2011

GHF Staff:
John Hurd, Technical Conservation Consultant
[email protected]

Dan Thompson, Director, Global Projects & GHN

[email protected]

Project Director:
Prof. Dr. Klaus Schmidt
Deutsches Archaeologisches Institut
[email protected]
Site Background

The circa 150-meter-long and 300-meter-wide Early Neolithic site of Göbekli

Tepe (Belly Hill) is situated about 15 kilometers northeast of the modern city of
Şanlıurfa in southeastern Turkey at the highest point of an elongated mountain
range. Since 1995, archaeological work by the German Archaeological Institute
(DAI) has been carried out under the direction of Prof. Dr. Klaus Schmidt.
Göbekli Tepe is not a settlement but a sanctuary, probably a regional centre
where communities met to engage in complex rites. This interpretation is
supported by the fact that the annual excavation seasons since project
inception have provided no evidence of residential buildings or fortifications
but exclusively monumental, megalithic stone circles dating to the Pre-Pottery
Neolithic period (10th – 9th millennium B.C.). That the sophistication in
construction evidenced at the site existed in this period, one of the most
important in the history of humankind at the dawn of sedentary and farming
communities, makes Göbekli Tepe arguably one of the most important sites
under investigation today.

Monolithic T-shaped pillars, each often weighing several tons, were erected
here forming large enclosures in whose centre another pillar or pair of pillars
tower above all (see map of the enclosures and the pillars in situ). The arms
and hands depicted on their sides allow us to understand them as
anthropomorphic. Furthermore, there are large-sized reliefs added to these
pillars showing wild animals like large cats, bulls, wild boars, foxes, birds,
scorpions, spiders and snakes. These reliefs open the view upon a new and
unique picture language not known before, and the interpretation of these
remains an important factor of research discussion.

Professor Schmidt estimates that the life of the site may have been
approximately 1500 years and that during this time the tendency to build
enclosures containing T-shaped and relief-carved stones, and further ringed
with T-shaped stones, started with the construction of large examples and that
these reduced in size over time. It is clear to John Hurd that there is a
significant difference in sophistication of construction technology of the
mortars over this time.

As Professor Schmidt points out, the earliest part of the Göbekli Tepe site may
represent the peak of Stone Age technology which then declines towards the
end of the site’s life as the introduction of agricultural activity develops. This
would mean that Göbekli Tepe is not the start of a stage of human
development but in fact the end, implying that these advanced stone
technologies started at a much earlier time.

2
Above: Plan of principle excavations, 1995-2009.

3
Technical Background

Note that this report is concerned first and foremost with initial considerations
of site conservation, one of the four tenets of Preservation by Design®. Details
on planning, community development and partnerships – the other pillars of
Preservation by Design – are addressed elsewhere, most notably in the project
prospectus documentation or in reports devoted specifically to those topics.

Mortars.
Mortars are used in various forms and applications across the site.

While the large T-shaped stones seem to stand free and are occasionally
socketed into shallow sockets in the bedrock, occasionally consolidated by flint
and chert wedges or chinking stones, between the monoliths are mortared,
rough stone walls which support the monoliths in several techniques and also
divide spaces within the patterns of the position of monoliths in each
enclosure.

The rough stones are to some extent shaped on the face, and occasionally
there are protruding stones, whose function is not yet fully understood.

These walls vary enormously and are constructed at varying levels, may be
straight or concave, can support horizontal stones as tables or other flat
features, and the stones used may vary in size and so on, leading to the
conclusion that the walls represent a range of construction dates through the
life of the site. This opinion is further reinforced by the nature and content of
the mortars that support the wall.

The mortars across the site vary in colour, granulometry and composition and
this may be interpreted in several ways. In some positions the walls act as
buttressing walls against the hillside slope, and in these the “strength” of the
mortar, especially the clay proportion, may reflect this.

In another scenario the change in mortars my represent improvement or

decline of technology over time, and if this is the case then the mortars may
assist in providing chronological information for the construction of the walls
and the chronology of changes in the form of the enclosures.

Over the history of the site’s use there is compelling evidence that the
enclosures were intentionally backfilled and this backfill material is by no
means random or primitive, having evidence of clear lamina and also the use of
intentionally placed mortars.

4
Mortars throughout the site propose an extraordinary level of sophistication in
the understanding of their use. The attachment of mortar to stone is generally
significantly well achieved.

Generally when mortar is used in any masonry structure, the observer would
expect to see micro cracks and cleavage features related to shrinkage and
mechanical or vibrational activity over time. At Göbekli Tepe the attachment
between mortar and stone remains intimately close and strong. This implies
that construction was sophisticated enough to support the theory that the
stones were well wetted down during construction. The author regards this
attachment remarkable, given that these joints are 12,000 years old and that
an enormous effort needs to be made to transport water to mix the mortar and
to reduce suction by wetting the stones during the building process.

It is clear that careful characterisation and analysis will be required in order to

further understand the data and archaeological information that the mortars
offer.

Throughout the mortar series there is variable weathering to the level of the
mortars between the stones, from wall to wall. This is almost certainly a result
of the different seasons of excavation, with deeper erosion in those walls that
were uncovered earlier, or it may reflect differences in the mortar mixtures;
this needs careful study.

There is frequent evidence of insect activity within the mortars, especially

masonry bee holes, since excavation.

5
Above: Wall bay between monoliths.

Above: Detail of wall with good pointing, recently excavated.

6
Above: Detail of wall with eroded pointing, excavated in 2005.

7
Above: Detail of intentional infill, section.

8
Plasters.
In various positions of the walls there is occasional evidence of the mortars
extending to use of plasters. Those examined by the author were
unsophisticated plasters roughly applied by hand and “daubed” unevenly over
the wall. There are small marks in the daubed surface that the author
interprets as vestigial remains of small soft finger marks.

Above: Daubed plaster lower wall with damage to wall above.

9
Above: Detail of plaster with finger marks to left of scale.

Artificial Stone.
There are several very significant materials associated with the site, and these
include artificial stone (terrazzo) floors in some of the later enclosures and the
presence of examples of a “concrete” manufacture within re-used basins left
by the extraction of large stones from nearby quarries.

While there is no evidence of lime burning technology, there exists the strong
possibility of soft marly limestone minerals (Marl is normally in the index 5-15%
clays and 85-95% carbonates), in the climatic conditions, contemporary to
construction, of marshy ground in the forested plains below. These soft marls
when mixed with sands and aggregates would produce a type of concrete on
drying, and this may have been the source of the terrazzo floors. Clearly more
research is needed into this possibility of the use of concretes at such an early
date.

10
Above: “Terrazzo” floor in need of consolidation.

Above: Monolith excavation pit in quarry, reused and containing traces of “concrete”.

11
Stone.
Göbekli Tepe lies in a limestone region, and around the limestone monuments
in the circles of the archaeological site is a landscape profoundly changed by
the working of limestone, stripping the surface off the natural limestone
pavements, with wasted limestone rocks lying on the surface. Klaus Schmidt
points out that this is not a natural landscape but a cultural landscape
converted through human activity.

Within the landscape are numerous piles of knapped flint which demonstrate a
continual flint knapping industry. Flint tools and other stone axes were used to
extract the huge T-shaped monoliths, and although not immediately available
on the hill at the site, it is not in short supply.

There is no evidence of burning as part of the quarrying activity, but the supply
of water artificially brought to the quarries and to the Göbekli hill is strongly in
evidence.

The monoliths themselves show a very fine ashlar surface, decorated with
representative and in some cases contiguous bas relief. This is not primitive
work and implies an historic and well experienced stone technology.

The site team includes a stone conservator/mason, Andreas Gotz,

[email protected] who is an active and capable conservator. Together
with Eduard Knoll, [email protected] architect and civil engineer,
they represent the core stone team.

12
Above: Monoliths in Enclosure B showing reuse/repair.

13
Above: Detail of fragile ancient repair, also in Enclosure B.

14
Above: Bas relief decoration depicting a fox.

15
Above: A contemporary patch in stone floor with basalt wedging stones.

16
State of conservation

Until now little formal conservation has been achieved, but the archaeological
team has taken steps to preserve the mortared and plaster-daubed walls by
introducing an outer dry stone protective lamina which is providing very
efficient temporary protection to the ancient walls.

Clearly this cannot remain in place as the outer lamina is somewhat unsightly
and alters the dimensions of the main structures.

Above the upper smaller enclosures is a simple and functional metal shelter
structure, which is serviceable.

Some limited stone conservation has been achieved with unset monoliths
supported by raking shores and anchored cables. One large broken stone has
been rejoined using an epoxy resin adhesive. This repair urgently requires
further attention in order to prevent free moisture from entering at the edges
of the repair during winter 2011, and John Hurd recommends that a temporary
hydrophilic plaster be applied across the crack to control water ingress that
may contribute to hydrolysis and swelling of the adhesive over time.

While archaeological research has gone ahead for many years, the conservation
of what has been excavated has not kept pace, and there is now a considerable
backlog of work needed both in repair and reassembly of broken stones and in
the conservation of mortars, plasters and extraordinary “terrazzo” floor
coverings.

Andreas Gotz and Eduard Knoll have a continual programme of stone repair,
and it is important that they start training local people in the techniques used
for stone repair or the conservation will continually fall behind the
archaeological research.

John Hurd is happy to lead the conservation of the earth mortars and plasters,
should the project director require this, and this also will need to produce local
conservators to continue the work during the whole season.

During the spring season of 2012, both the stone conservators and the earth
conservators should conduct intensive training courses to ensure that work can
continue for the whole season.

17
Conservation requirements

An urgent conservation requirement is the need to gather pertinent data in

order to start the process of understanding risks to the site and its
archaeological features.

The first requirement is the establishment of weather stations in at least two

contexts: on the open hill and in the deepest excavation trenches.

These need to monitor all normal climatic measurements, and Hurd proposes
that these store data in electronic memories which can be transferred from
time to time by USB connections into team computers. A member of the
archaeological team will need to be identified to co-ordinate this activity, and
the data must then be used to interpret and define the risks to the site.

The other principal conservation requirements centre around the stone,

mortars, plaster and terrazzos in equal importance, with monoliths occasionally
damaged, and historic evidence of support of broken elements employing
earth-mortared support walls as collared buttresses and so on.

The conservation challenge mainly addressed by the author is the consolidation

and protection of the earth mortars, plasters and artificial stone floors.

The mortars and plasters are very vulnerable to weathering and erosion
through the action of wind and rain and especially in the rare event of a violent
rainstorm, which would, in a short period, do widespread damage. Climate
change must be engaged as a serious risk at Göbekli Tepe.

Shelter structure

As a first principle, there is no question that the most important conservation

tool for this important site which remains at high risk from environmental
erosion is the construction and use of shelter structures, and to that end a very
functional shelter structure had been designed to shelter the main excavated
monumental enclosures.

However, this design has now been superseded by an alternative design

supplied by the German Archaeological Institute, and Hurd has concerns with
this structure.

The Archaeological team have limited options for placing the support posts for
any structure. The new design seems to ignore the ideal positions proposed by
Klaus Schmidt and places the posts in other positions, suggesting that the new
design was achieved for form and beauty rather than function.

18
A shelter structure is a conservation tool, and function should always precede
form. No useful design could be made without the designer having intimate
knowledge of the function.

The new design foresees a tensile textile as the main protective roof, and this
is very acceptable within the constraints of the design life of the structure and
the need for diligent maintenance.

The environmental and climatic conditions at Göbekli Tepe include lively and
dusty winds that will deposit large quantities of dust on the shelter roof.
Design of the roof must therefore reflect this condition and as such should
avoid valleys and other features within which the dust may settle. When dust
settles on the roof and other elements of the structure, then during rainfall the
dust will be deposited on the archaeology surrounding the roof, and drainage
systems may become blocked with potentially disastrous impact on the
moisture content of the archaeology. Dust, forming on the proposed white
tensile roof, will quickly make the roof unsightly and inefficient. John Hurd
urges the German Archaeological Institute to revisit the design with these
challenges in mind.

Hurd stresses that the chemical integrity of the Göbekli Tepe site must be a
priority, and this includes the protection of the site from pollutants such as the
introduction of concretes made with ordinary Portland cements and other
chemical admixtures.

The foundations for posts of any shelter structures should be isolated from the
archaeology by placing an effective separation layer between the concrete
foundations and the archaeology. This separation layer must perform as a
“cordon sanitaire” between introduced materials and the original materials on
the site.

The research techniques of archaeology are changing in terms of many micro

research techniques which can gather important chemical data from the site
and this must be protected in so far as possible.

Conservation Materials Research and Testing

Minimal Field Laboratory.

As conservation progresses, supported through Global Heritage Fund, there will
be a need to establish a small field laboratory on site.

While initial research towards characterising the mortars and plasters will be
achieved by outside laboratories, conservation on site will greatly benefit from
basic characterisation equipment, a dry sieving tower, sedimentation vessels
and a small binocular photo microscope.

19
Tools, trowels, pointing irons, water spraying equipment and so forth will also
be required as intervention starts. Hurd will assemble a complete list of
conservation equipment required for the 2012 Spring season.

Permissions and Consents.

Since permissions and consents will be required for the sampling of materials to
be examined, characterised and tested, Hurd proposes that this be done, as far
as possible, by Turkish institutes.

There are two main institutes that may be acceptable partners in this activity.
The Middle Eastern technical University (METU) in Ankara and the local Harran
University Department of Agriculture in Şanlıurfa.

Mortars and Plasters.

Mortars, plasters and artificial stone floors will need to be characterised,
especially through the production of granulometry data to allow for comparison
of both chronological and reuse purposes. This is a relatively simple process
and can be carried out in any well equipped soil science laboratory.

Microscopy.
Samples can be examined on site through the use of a small standard binocular
photo microscope; this can have many uses for both the recognition of
anomalous contents of building materials for conservation purposes and for
archaeological research examination of the same materials.

Mass Spectrometry.
In some cases, materials samples may require a deeper understanding,
especially when specific clays or general admixtures to materials require
further characterisation. This can be achieved through mass spectrometry
which is available at Harran University in Şanlıurfa or through other
spectrometry techniques available at METU.

Test Walls.
Once conservation materials have been identified for use on the monument,
these should be tested for at least one full year to examine their performance.
Tests cannot be ethically carried out on the authentic and ancient material and
so Hurd proposes that a series of small test walls be constructed on site for the
purpose of experimentation.

20
Proposed interventions

Temporary Protection.
Before the end on the 2011 season, temporary protection must be made of
certain vulnerable aspects of the site. These objects are vulnerable to climate
erosion, mechanical damage and theft.

Ed Knoll is listing and prioritising these temporary protection incidences and

methods now and in co-operation with the archaeological team and Hurd and
the conservation team.

Active Conservation Intervention:

Stone.
In respect of the stone conservation and following discussion with Andreas Gotz
and Eduard Knoll, Hurd has proposed that in depth discussion and research is
made into the selection of appropriate adhesives for the limestone repairs to
be made at Göbekli Tepe. Hurd will introduce Andreas Gotz to the ICOMOS
International Scientific Committee on the Conservation of Stone, and Gotz may
use this committee as a forum to compare advice from stone conservators
around the world, prior to deeper discussion with adhesive manufacturers, in
respect of stone type, climatic conditions and so on.

Mortars and Plasters.

In respect of the conservation of mortars, after the materials have been
characterised, a series of tests will be conducted.

It is very likely that new material, similar or identical in composition to that

found in the historic context, will be added in a “repointing” process.

As this process continues, it will be necessary to be able to clearly identify the

difference between authentic original material and conservation materials.

Since clear separation layers cannot be introduced for many reasons, especially
the need to avoid creating cleavage planes, the conservation material should
have an indicator medium introduced into the mix.

In this case Hurd recommends the addition of glass microspheres to the

conservation material to perform as an indicator and to add to the compressive
strength of the material.

Since the conservation repointing will require periodic maintenance, Hurd also
recommends that carbon or synthetic polymer rods be inserted horizontally
into the repointing mixes. As the conservation repointing is eroded in natural

21
conditions, these maintenance indicators will be revealed and will indicate that
further maintenance repointing is required.

Details of these ideas will evolve in detail as the conservation planning

becomes more detailed.

Artificial Stone Floors.

The artificial stone, “terrazzo” floors are vulnerable to climatic erosion,
mechanical erosion through wind, animals, flora and so on.

These terrazzos may require consolidation through repair to lacunae.

Detailed planning in this respect will follow as a greater understanding of these

floors emerge, through analysis.

Monitoring

Clearly all results of conservation intervention will need to be included in a

programme of inspection and monitoring. As mentioned above, in some cases
monitoring indicators can be used, but in general the appropriate monitoring
programme should be conducted by an experienced conservator.

Pre- and Post Shelter Structure Conditions.

Over the next years, environmental conditions will be subject to considerable
change, especially through the introduction of a shelter structure or structures.

The conservation programme will need to be designed with this in mind and
testing will need to reflect the change.

Training

Hurd sees the need for training to be given starting in the spring season of 2012
and this training will operate on two levels:

1. Internship and academic placements Schmidt may choose to encourage

conservation interns from Turkey and other places.

2. Local workers will require training to perform general conservation

activities during the intervention phases, and these would be selected for
comprehension and study.

22
The training of local archaeological workers to a higher level towards
conservation skills can be an important part of the GHF Preservation by Design
model for community development.

Team and Partners

GHF is partnering with the German Archaeological Institute (DAI) and the
Turkish Ministry of Culture and Tourism to preserve Göbekli Tepe.

The final composition of the conservation team is to be discussed further.

Material Stocks

There are several considerations in this area, and perhaps the main need is to
locate and collect appropriate earth materials in the surrounding landscape to
act as the basic content of the conservation intervention material for the
mortars, plasters and artificial stone floors. The Harran University earth
science faculty may be able to help in this respect.

Other admixtures, glass microspheres, possibly hydrated lime and so on will

need to be sourced from local suppliers, if possible, and found safe storage at
the dig house in Şanlıurfa.

23