New Israeli computer tech puts the pieces together for pottery archaeologists - Archaeology Israel News | Haaretz
New Israeli computer tech puts the pieces together for pottery archaeologists
Scholars not only get a 3-D picture containing tiny details, they get a mock-up of the whole vessel and the family it belongs to.
By Nir Hasson | Jan. 31, 2014 | 4:39 PM
Avshalom Karasik. He developed the program for his PhD with an expert in theoretical physics. Photo by Emil Salman
Sometimes sherds create a whole vase. Photo by Emil Salman
In the Babylonian Talmud, Tractate Shabbat, Rabbi Jose praises the vessels of Kfar Hananya, deeming them “not likely to burst. ” Sure enough, excavations near Hananya Junction in the north have uncovered a village of potters from the Roman era. This includes two ovens 200 meters apart whose products seemed so similar that archaeologists thought they may have belonged to the same artisan.
But after the Kfar Hananya sherds were digitally scanned by a unique new system and the data were fed into a computer, it was clear these were two different workshops, two different generations and maybe even two different periods. The computer system, now being used by the Israel Antiquities Authority, heralds a new era for the archaeological study of pottery and similar finds.
“This technology opens up an entirely new set of questions that until now we couldn’t have asked,” says Prof. Gideon Avni, head of surveys and excavations at the Israel Antiquities Authority.
Sherds are the most common archaeological find in Israel – these broken pieces of pottery are found each year in the hundreds of thousands or millions. But then comes the really hard work – documenting them, drawing them, trying to imagine the vessel they came from, and classifying the sherd based on vessels from other sites.
The final goal of the scanner and accompanying software, which were developed by the antiquities authority, the Weizmann Institute of Science and the Hebrew University, is to establish a national database of sherds, a kind of sherd Google.
Scholars who scan their findings not only get a 3-D picture of a sherd in all its tiny details, they get a mock-up of the whole vessel and the family of vessels it belongs to. They can sometimes even guess at the potter who worked on it 1,000, 2,000 or 5,000 years ago. The more sherds fed into the program, the more connections to be made between sites and cultures.
The program was developed by Dr. Avshalom Karasik for his PhD with Prof. Uzy Smilansky of the Weizmann Institute, an expert in theoretical physics. “I studied archaeology and mathematics because those were the two fields that interested me,” Karasik says. “I never thought the two could be integrated.”
Archaeologists have been using sherds since the 19th century to date and classify material culture. The digital scanning of clay vessels isn’t new; archaeologists have been using scanners to produce high-level photos since the 1990s.
But Karasik’s algorithms can turn 3-D pictures into a research tool. Since most vessels were produced on pottery wheels, each has an axis around which it was created. Finding this axis makes it possible to recreate the entire vessel.
“The computer knows how the fragment stood in space,” Karasik says. “Next it draws the vessel’s profile, and it knows how to compare it with other vessels and find the potter’s ‘signature,’ the final punch he gave the vessel.”
The antiquities authority plans to use the system regularly. According to Dr. Zvi Greenhut, head of the authority’s laboratory, this will be the “ultimate information center for archaeologists.”
Tens of thousands of sherds have been scanned into the database, so now the system is trying to resolve a few controversies. For example, excavations at Tel Hazor in the north have revealed clay vessels similar to those found over the border at Tyre, Lebanon. Archaeologists at both Hazor and Tyre have no doubt that their sites were connected in the Iron Age 3,000 years ago.
But each side claims that the vessels were manufactured at its site. Israeli scholars can’t scan the vessels excavated in Lebanon, but they have scanned drawings of them and compared them with those from Hazor. The result – the two groups are clearly different.
Despite the system’s capabilities, the antiquities authority’s Avni says human archaeologists will always play a role.
“Some vessels are very similar; for example, those of the Hellenistic and the Middle Bronze Age,” Avni says, citing periods 3,000 years apart. “The program will think they’re from the same period. That’s where people have to use their heads.”