When we first received the permit to sift through the Temple Mount soil, we did not know how to approach the vast amount of debris, nor how many volunteers would be needed for such a task. The soil from the Temple Mount contains large amounts of ash and dust which coat the natural stones and archaeological finds. This makes it very difficult to differentiate between them. We discovered that using water in the sifting process was more effective in helping us uncover the archaeological material. Using this new system, we quickly began finding a large variety of artifacts, such as jewelry, ancient warfare implements, bone tools, coins, and fragments of architectural members. Gradually. We gradually developed the wet sifting technique and constructed a large green house equipped with dozens of sifting stands with hoses. This methodology has been successfully used at our site and has been adopted by other excavation projects in Jerusalem and Israel with similar needs. In some cases these other excavations prefer to hire our sifting services, and deliver their soil to our sifting facility. The number of seals and seal impressions found in Jerusalem has increased dramatically since the wet sifting method was introduced by our project.

Wet Sifting Process

אורן כהן1

The soil is first taken from labeled heaps of earth corresponding to the different areas on the Temple Mount from which they were taken. To reduce the amount of time and water necessary for dissolving the mud adhering to the finds, the soil is first dry sifted (using a special device that we constructed),  and separated  into buckets.


The buckets are then brought into our greenhouse and filled with water to soak the soil and loosen the dirt from the stones and artifacts. This makes the washing on the screens more rapid and efficient.

The buckets are then wet-sifted by staff or volunteers at the sifting stands using the spray taps. The volunteers scrutinize the washed material and collect the artifacts in cups by category. The artifacts are sorted into six main categories: Pottery, Glass, Bones and Shell, Mosaic Tessera (cubes), Metal, and Special Stones and Plaster



Every screen is checked by our staff members before the artifacts are deposited from the cups into colored buckets in the center of the greenhouse. The leftover stones and non-archaeological material is discarded. Special finds, like coins, are labeled on the spot and kept in a safe place on the archaeologist’s table.


After sifting, one of the archaeologists on staff shows the volunteers what they have found, what it means, and why it is important. They compare the artifacts that were found by the group to the artifacts in our display table.  The archaeologist tries to put the day’s work into context so that the volunteers can understand exactly what we are doing and the purpose of our project.

At the end of the day, the material in the colored buckets is sorted, labeled, and cleaned in order to be further studied by the site’s archaeologist the following day. All archaeological material is counted and put into our statistical database. Indicative or special finds are then brought into our laboratory to be studied and classified more extensively by experts in specific time periods or materials. Finds are also photographed, drawn, and readied for publication. After then classification and typology process, the data is used for advanced statistical analysis and data mining to help us reveal the hidden distribution patterns buried within the huge database, which eventually will help us understand the full significance of our finds.

Archaeological Research with Finds that are Out-of-Context

When an archaeological site is excavated brutally with heavy machinery, ignoring the stratigraphy, context, and finds, major archaeological data is lost and can never be retrieved. However, out-of-context finds that are recovered from archaeological sites still retain some valuable archaeological information. These finds, when studied thoroughly, can increase the information about the site, especially if it hasn’t previously been excavated. Such is the case of the Temple Mount.

Comparative Analysis with other Sites in Jerusalem

Most of our finds can be identified and dated by typologically matching them to parallel artifacts which were found in clear contexts elsewhere. This methodology is widely used in archaeological surveys, which study sites only by gathering artifacts from the topsoil. In surveys, the finds collected are considered to be a reliable sample of the common types of finds in the stratigraphic layers below. Eventually, after sorting the finds, general conclusions may be reached regarding the size and occupation of the site throughout history. Usually, this kind of research uses only small samples, and does not focus on the material culture characteristics of the site. In the case of the Temple Mount debris, since the sample size is enormous, the statistical analysis capabilities are much wider. This is a good thing. By the time we complete the sorting and classification of the finds, the data recorded will not only teach us about the distribution and the intensity of activity in different time periods, but will also provide information regarding the distribution of the various classes and types of finds. Subsequently, types of finds can be studied more comprehensively by comparing the distribution to other sites in Jerusalem.

Control GroupsFor this purpose, we sampled soil from several sites on the slopes of Jerusalem’s Old City and sifted it at the Sifting Project facility in order to create statistical control groups that can then be compared with the Temple Mount finds. The samples from these sites contained small and mixed finds from various periods, which is similar to what we have found in the soil from the Temple Mount. We have sampled eight such sites, and will consider sampling more in the future if necessary.

Clustering Types of Prevalent Finds

We can apply further quantitative analysis to the comparative study mentioned above, as well as to the distribution of finds in the Temple Mount debris. It appears that there are differences between the distributions of finds in different sections of the dump. We labeled the different heaps in the dump and thus succeeded in differentiating between different areas on the Temple Mount. There are major differences in the ratios of artifacts from different time periods between area T, the dumps that were kept in a downtown compound, and area K, the major dump at the Kidron valley. There are also clear differences among sub-sections of area K.

Location of Debris

Original Location of the Debris on the Temple Mount

From the early stages of the Sifting Project, we noticed that there were differences in the frequency of certain types of finds from different areas. For example, one heap might have more metal artifacts while another had a larger percentage of mosaic tesserae. Moreover, similar finds, or fragments of the same object, were often discovered within a short period of time. This suggests that these objects were originally next to each other and that the mixing of the layers was less complete. It seems as if the removal of the earth created clusters of artifacts from the same in situ context in the earth heaps that we are now sifting.

The full significance of labeling and dividing the dump material into different areas has only recently been realized. We have discovered that artifacts which we assume to be from the same context are also distributed among the heaps in a similar manner. For example, we found that artifacts associated with the horses of the Crusader Era and Knights Templar are distributed in a similar way among the dump sections.

Consequently, we can define a statistical distribution “fingerprint” for each prevalent artifact type.  Artifacts with a similar “fingerprint” could have originated from the same in situ context before the Waqf dig. Currently, we are still investigating the application and implications of this statistical method. Only after completing the classification and sorting process will we be able to create a full data table adequate for such an analysis. With the help of data mining techniques (such as those used by cutting edge software like SAS Enterprise Miner) we will be capable of fully estimating the value and proficiency of this method. If indeed we are able to retrieve valuable information through this analysis, it will be a substantial innovation in archaeological methodology and theory which could then be applied by other archaeologists who focus their research on the excavation of fills or site surveys.

Data mining

Data Mining our Research

%d bloggers like this: